http://wikis.swarthmore.edu/ling073/api.php?action=feedcontributions&feedformat=atom&user=Slim1
LING073 - User contributions [en]
2026-10-25T16:02:28Z
User contributions
MediaWiki 1.27.7
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5418
User:Slim1/Final project
2017-05-11T04:45:30Z
<p>Slim1: </p>
<hr />
<div>=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Description=<br />
==Approach==<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
I used [http://scripts.sil.org/ukelele SIL's Ukelele] to create the MacOS keyboard.<br />
<br />
I used the [https://msdn.microsoft.com/en-us/globalization/keyboardlayouts.aspx Microsoft Keyboard Layout Creator] to create the Windows keyboard.<br />
<br />
I created and edited an xkb file in Linux to create the Linux keyboard.<br />
<br />
==Design Principles==<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
[[File:Kalmyk (phonetic) layout.png|thumb|center|The Linux phonetic layout I created]]<br />
[[File:Xalphon firstlevel.png|thumb|center|First level of the MacOS phonetic layout I created]]<br />
[[File:Xalphon thirdlevel.png|thumb|center|Third level of the MacOS phonetic layout I created]]<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
[[File:Xal firstlevel.png|thumb|center|First level of the MacOS layout I created]]<br />
[[File:Xal thirdlevel.png|thumb|center|Third level of the MacOS layout I created]]<br />
<br />
<br />
=Evaluation=<br />
I evaluated the keyboard on myself. This is obviously not a fair measure of the utility of the keyboard for various reasons, and only speakers of the language can truly evaluate the utility of the Tatar-Russian based layout. However, I provide parameters and results so they may be used as a source of comparison.<br />
<br />
The "test" consisted of measuring time, number of referrals to the keyboard layout display, and number of typos while typing a random corpus of 50 words that I had not seen before. I ran three trials and averaged the results. I did not correct typos as I typed. I used typing in English as the control, since I speak it fluently and I am also very familiar with the layout, which is the most common state of typists of a language.<br />
<br />
; English layout ("control group")<br />
:* time spent typing (seconds): 47.3<br />
:* number of referrals to the layout display: 0<br />
:* typos (number of words typed incorrectly): 2.3<br />
<br />
; Phonetic layout<br />
: I used a prototype version of the phonetic layout - a layout I created when the only exposure I had to the language and orthography was looking up the alphabet - throughout the semester. Thus, it did not take me long to become accustomed to this new phonetic version. Some variables to note: I do not speak the language, but am reasonably familiar with the orthography; I am the designer of the keyboard. <br />
:* time spent typing (seconds): 109.4<br />
:* number of referrals to the layout display: 2.6<br />
:* typos (number of words typed incorrectly): 7.0<br />
<br />
; Tatar-based layout<br />
: The three experiments include my first usage of the keyboard. Some variables to note: I do not speak the language; I am the designer of the keyboard.<br />
:* time spent typing (seconds): 261.8<br />
:* number of referrals to the layout display: had it displayed constantly<br />
:* typos (number of words typed incorrectly): 3.0<br />
: Since I don't know the layout, I had it displayed to the side as I typed. The low number of typos is due to the slow speed and frequent referrals to the corpus I was typing.<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5417
User:Slim1/Final project
2017-05-11T04:28:54Z
<p>Slim1: </p>
<hr />
<div>=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Description=<br />
==Approach==<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
I used [http://scripts.sil.org/ukelele SIL's Ukelele] to create the MacOS keyboard.<br />
<br />
I used the [https://msdn.microsoft.com/en-us/globalization/keyboardlayouts.aspx Microsoft Keyboard Layout Creator] to create the Windows keyboard.<br />
<br />
I created and edited an xkb file in Linux to create the Linux keyboard.<br />
<br />
==Design Principles==<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
[[File:Kalmyk (phonetic) layout.png|thumb|center|The Linux phonetic layout I created]]<br />
[[File:Xalphon firstlevel.png|thumb|center|First level of the MacOS phonetic layout I created]]<br />
[[File:Xalphon thirdlevel.png|thumb|center|Third level of the MacOS phonetic layout I created]]<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
[[File:Xal firstlevel.png|thumb|center|First level of the MacOS layout I created]]<br />
[[File:Xal thirdlevel.png|thumb|center|Third level of the MacOS layout I created]]<br />
<br />
<br />
=Evaluation=<br />
; Phonetic keyboard<br />
: I used a prototype version of the phonetic keyboard - a layout I created when the only exposure I had to the language and orthography was looking up the alphabet - throughout the semester. Thus, it did not take me long to become accustomed to this new phonetic version. Some variables to note: I do not speak the language, but am reasonably familiar with the orthography; I am the designer of the keyboard. In typing a random section of corpus of 50 words, the average across three tries were:<br />
:* time spent typing: 109.4 sec<br />
:* number of referrals to the layout display: 2.8<br />
:* typos (number of words typed incorrectly): 7.6<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5416
User:Slim1/Final project
2017-05-11T04:15:01Z
<p>Slim1: </p>
<hr />
<div>=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Description=<br />
==Approach==<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
I used [http://scripts.sil.org/ukelele SIL's Ukelele] to create the MacOS keyboard.<br />
<br />
I used the [https://msdn.microsoft.com/en-us/globalization/keyboardlayouts.aspx Microsoft Keyboard Layout Creator] to create the Windows keyboard.<br />
<br />
I created and edited an xkb file in Linux to create the Linux keyboard.<br />
<br />
==Design Principles==<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
[[File:Kalmyk (phonetic) layout.png|thumb|center|The Linux phonetic layout I created]]<br />
[[File:Xalphon firstlevel.png|thumb|center|First level of the MacOS phonetic layout I created]]<br />
[[File:Xalphon thirdlevel.png|thumb|center|Third level of the MacOS phonetic layout I created]]<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
[[File:Xal firstlevel.png|thumb|center|First level of the MacOS layout I created]]<br />
[[File:Xal thirdlevel.png|thumb|center|Third level of the MacOS layout I created]]<br />
<br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=File:Xal_thirdlevel.png&diff=5415
File:Xal thirdlevel.png
2017-05-11T04:14:45Z
<p>Slim1: </p>
<hr />
<div></div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=File:Xalphon_thirdlevel.png&diff=5414
File:Xalphon thirdlevel.png
2017-05-11T04:14:10Z
<p>Slim1: </p>
<hr />
<div></div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5413
User:Slim1/Final project
2017-05-11T04:12:24Z
<p>Slim1: </p>
<hr />
<div>=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Description=<br />
==Approach==<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
I used [http://scripts.sil.org/ukelele SIL's Ukelele] to create the MacOS keyboard.<br />
<br />
I used the [https://msdn.microsoft.com/en-us/globalization/keyboardlayouts.aspx Microsoft Keyboard Layout Creator] to create the Windows keyboard.<br />
<br />
I created and edited an xkb file in Linux to create the Linux keyboard.<br />
<br />
==Design Principles==<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
[[File:Kalmyk (phonetic) layout.png|thumb|center|The Linux phonetic layout I created]]<br />
[[File:Xalphon firstlevel.png|thumb|center|First level of the MacOS phonetic layout I created]]<br />
[[File:Xalphon secondlevel.png|thumb|center|First level of the MacOS phonetic layout I created]]<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
[[File:Xal firstlevel.png|thumb|center|First level of the MacOS layout I created]]<br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=File:Xal_firstlevel.png&diff=5412
File:Xal firstlevel.png
2017-05-11T04:11:27Z
<p>Slim1: </p>
<hr />
<div></div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=File:Xalphon_secondlevel.png&diff=5411
File:Xalphon secondlevel.png
2017-05-11T04:10:44Z
<p>Slim1: </p>
<hr />
<div></div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=File:Xalphon_firstlevel.png&diff=5410
File:Xalphon firstlevel.png
2017-05-11T04:08:36Z
<p>Slim1: </p>
<hr />
<div></div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5397
User:Slim1/Final project
2017-05-11T02:53:48Z
<p>Slim1: </p>
<hr />
<div>=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Description=<br />
==Approach==<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
I used [http://scripts.sil.org/ukelele SIL's Ukelele] to create the MacOS keyboard.<br />
<br />
I used the [https://msdn.microsoft.com/en-us/globalization/keyboardlayouts.aspx Microsoft Keyboard Layout Creator] to create the Windows keyboard.<br />
<br />
I created and edited an xkb file in Linux to create the Linux keyboard.<br />
<br />
==Design Principles==<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
[[File:Kalmyk (phonetic) layout.png|thumb|center|The Linux phonetic layout I created]]<br />
<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
<br />
<br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5327
User:Slim1/Final project
2017-05-10T10:52:14Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
[[File:Kalmyk (phonetic) layout.png|thumb|center|The Linux phonetic layout I created]]<br />
<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5326
User:Slim1/Final project
2017-05-10T10:47:54Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
[[File:Kalmyk (phonetic) layout.png|thumb|center|The Linux phonetic layout I created]]<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5325
User:Slim1/Final project
2017-05-10T10:46:54Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
[[Kalmyk (phonetic) layout.png|thumb|center|The phonetic layout I created]]<br />
<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=File:Kalmyk_(phonetic)_layout.png&diff=5324
File:Kalmyk (phonetic) layout.png
2017-05-10T10:46:08Z
<p>Slim1: </p>
<hr />
<div></div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5323
User:Slim1/Final project
2017-05-10T10:44:54Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
[[File:Russian.png|thumb|center|Linux Russian layout]]<br />
[[File:Russian.png|thumb|center|Linux Tatar layout]]<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=File:Tatar.png&diff=5322
File:Tatar.png
2017-05-10T10:44:24Z
<p>Slim1: </p>
<hr />
<div></div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=File:Russian.png&diff=5321
File:Russian.png
2017-05-10T10:43:40Z
<p>Slim1: </p>
<hr />
<div></div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5320
User:Slim1/Final project
2017-05-10T10:43:04Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|thumb|center|Linux Kalmyk layout]]<br />
<br />
<br />
<br />
<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5319
User:Slim1/Final project
2017-05-10T10:42:49Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
[[File:Kalmyk unix.png|200px|center|Linux Kalmyk layout]]<br />
<br />
<br />
<br />
<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5318
User:Slim1/Final project
2017-05-10T10:36:31Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a English-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Tatar-based layout.<br />
<br />
The English phonetic layout is for non-native speakers and learners of Kalmyk who are unfamiliar with the Russian keyboard layout. For the phonetic layouts, I tried to stay as consistent with the English keyboard layout as possible, mapping the Cyrillic characters to the position of the equivalent-sounding English characters. For Cyrillic characters without an English counterpart, I either placed them near the closest-sounding English keys, or placed them near related Cyrillic characters. I also tried to preserve the location of important punctuation keys in the English layout.<br />
<br />
The Tatar-based layout is for Kalmyk speakers with little knowledge of English. The official xkb Kalmyk layout in Linux is equal to the Russian layout, with the six additional letters in Kalmyk added on the number row. This is convenient for Kalmyk speakers who frequently type in Russian, as it allows the user to type in both languages without switching keyboards. However, this is also inconvenient because the six letters that are not present in Russian are used quite frequently in Kalmyk. The new layout attempted to remedy this problem. It was closely based on the Linux Tatar layout, as they use the equal set of characters. The Tatar layout also does not differ too much from the Russian layout, which will make it easier for people who are used to the Russian keyboard to learn this layout. Once again, I placed important punctuations in accessible locations.<br />
<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5112
User:Slim1/Final project
2017-05-07T00:49:09Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a Latin-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Russian-based layout.<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5111
User:Slim1/Final project
2017-05-07T00:48:51Z
<p>Slim1: </p>
<hr />
<div>I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a Latin-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Russian-based layout.<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5110
User:Slim1/Final project
2017-05-07T00:48:40Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
I chose the [https://wikis.swarthmore.edu/ling073/Final_projects/Keyboard_layouts keyboard layout] project. I developed installable Kalmyk keyboard layouts for three different operating systems: Linux, Windows, and MacOS. For Linux, I created a Latin-keyboard based phonetic layout; for Windows and MacOS, I created a phonetic layout and a Russian-based layout.<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5109
User:Slim1/Final project
2017-05-07T00:22:00Z
<p>Slim1: </p>
<hr />
<div>=Description=<br />
<br />
=Code=<br />
The code is available on my public [https://github.com/sjorcindy/xal-keyboards github repository]. <br />
<br />
=Evaluation=<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5036
User:Slim1/Final project
2017-05-05T01:50:39Z
<p>Slim1: </p>
<hr />
<div>template<br />
<br />
<br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=User:Slim1/Final_project&diff=5035
User:Slim1/Final project
2017-05-05T01:50:13Z
<p>Slim1: Created page with " Category:sp17_FinalProjects Category:Kalmyk"</p>
<hr />
<div><br />
[[Category:sp17_FinalProjects]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4971
Kalmyk/Universal Dependencies
2017-04-29T00:17:23Z
<p>Slim1: </p>
<hr />
<div>The developed models are in the <code>/dev</code> folder in the [https://github.swarthmore.edu/slim1/ling073-xal-corpus ling073-xal-corpus repository].<br />
<br />
=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In my corpora, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "the people came eastward to the field of Shinar." Here, "field" has the nmod relation to "Shinar."<br />
<pre><br />
"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people <br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east<br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction<br />
"<Шинарин>"<br />
"Шинар" np gen @obj #4->6 !Shinar<br />
"<талвңд>"<br />
"талвң" n sg dat @nmod #5->4 !field <br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came<br />
</pre><br />
The following phrase translates to "we must mix (confuse) their language." Here, "language" has the nmod relation to "their."<br />
<pre><br />
"<Бидн>"<br />
"бидн" prn pers p1 pl inc nom @nsubj #6->9 !we<br />
"<теднә>"<br />
"тедн" prn pers p3 pl gen @amod #7->8 !they<br />
"<келинь>"<br />
"кел" n px3sp @nmod #8->7 !language<br />
"<хутхулхмн>"<br />
"хутхулхмн" v @root #9->0 !mix<br />
</pre><br />
<br />
===amod===<br />
amod is an adjective or an adjectival phrase that modifies a noun. This is a fairly straightforward relation that came up many times exactly as such in my corpora. <br /><br />
The following phrase translates to "Kalmyk nation" (Kalmykia). "Kalmyk" has the amod relation to "nation."<br />
<pre><br />
"<Хальмг>"<br />
"хальмг" adj @amod #9->10 !Kalmyk<br />
"<улс>"<br />
"улс" n sg nom @nsubj #10->13 !nation<br />
</pre><br />
The following phrase translates to "too few present (day) youths." "too few" and "present" both have the amod relation to "youths."<br />
<pre><br />
"<хор-хомс>"<br />
"хор-хомс" adj @amod #11->13 !too few<br />
"<одаһин>"<br />
"одаһин" adj @amod #12->13 !present<br />
"<баһчуд>"<br />
"баһчуд" n sg nom @nsubj #13->16 !youth<br />
</pre><br />
<br />
===det===<br />
det is a determiner head of a nominal. As expected, the words tagged as <det> had this head.<br />
The following phrase translates to "this day."<br />
<pre><br />
"<Эн>"<br />
"эн" det dem @det #1->2 !this<br />
"<өдр>"<br />
"өдр" n sg nom @nsubj #2->9 !day<br />
</pre><br />
The following phrase translates to "many people."<br />
<pre><br />
"<кесг>"<br />
"кесг" det @det #7->8 !many<br />
"<күн>"<br />
"күн" n sg nom @nsubj #8->10 !people<br />
</pre><br />
<br />
===obj===<br />
obj is the second most core argument of a verb. In the corpora, the nominal near the verb usually had this head, since Kalmyk is a SOV language.<br />
The following phrase translates to "this is called Operation Ulussy." "Operation" has the obj relation to the verb.<br />
<pre><br />
"<Эн>"<br />
"эн" det dem @nsubj #1->6 !this<br />
"<«>"<br />
"«" lquot @punct #2->3<br />
"<Операц>"<br />
"Операц" n sg nom @obj #3->6 !Operation<br />
"<Улусы>"<br />
"Улусы" np sg nom @nmod #4->3 !Ulussy<br />
"<»>"<br />
"»" rquot @punct #5->3<br />
"<нерлх>"<br />
"нерлх" v inf @root #6->0 !is called<br />
</pre><br />
<br />
The following phrase translates to "used tar instead of mortar" (the order of "tar" and "mortar" is reversed). <br />
<pre><br />
"<урсхлын>"<br />
"урсхлын" n sg gen @obj #5->8 !mortar<br />
"<ормд>"<br />
"ормд" prep @advmod #6->7 !instead of<br />
"<салм>"<br />
"салм" n sg nom @nsubj #7->8 !tar<br />
"<олзлв>"<br />
"олзл" v pret @root #8->0 !used<br />
</pre><br />
<br />
===acl===<br />
acl is used for finite and non-finite clauses that modify a nominal. In my corpora, acl were clauses that modified a noun.<br />
The following phrase translates to "they found the baby, who was lying in the manger." The core of the clause "lying in the manger" has the acl relation to "baby."<br />
<pre><br />
"<отльгт>"<br />
"отльгт" n sg nom @iobj #12->13 !manger<br />
"<кевтсн>"<br />
"кевт" v part past @acl #13->15 !lying<br />
"<теднə>"<br />
"теднə" prn pers p3 pl @nsubj #14->16 !they<br />
"<үринь>"<br />
"үр" n px3sp @obj #15->16 !baby<br />
"<үзцхəв>"<br />
"үзцх" v pret @root #16->9 !found<br />
</pre><br />
<br />
The following phrase translates to "God saw the city and tower that the people were building." The core of the clause "that the people were building" has the acl relation to "city."<br />
<pre><br />
"<әмтнә>"<br />
"әмтнә" n sg @obl #1->3 !people<br />
"<тосхҗах>"<br />
"тосхҗах" v fut @acl #2->3 !build<br />
"<балһс>"<br />
"балһс" n sg nom @conj #3->5 !city<br />
"<болн>"<br />
"болн" conjcoo @cc #4->5 !and<br />
"<бәәшң>"<br />
"бәәшң" n sg nom @csubj #5->6 !tower<br />
"<хәләхәр>"<br />
"хәләхәр" v @ccomp #6->10 !look<br />
"<Деед>"<br />
"Деед" adj @amod #7->8 !higher<br />
"<Эзн>"<br />
"эзн" n sg nom @nsubj #8->12 !master<br />
</pre><br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk&diff=4970
Kalmyk
2017-04-29T00:15:22Z
<p>Slim1: /* Developed Resources */</p>
<hr />
<div>=Developed Resources=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk/Keyboard Kalmyk keyboard] <br /><br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Corpus (Github)] <br /><br />
[https://wikis.swarthmore.edu/ling073/Kalmyk/Grammar Grammar Documentation] <br /><br />
[https://wikis.swarthmore.edu/ling073/Kalmyk/Transducer Transducer] <br /><br />
[https://wikis.swarthmore.edu/ling073/Kalmyk/Disambiguation Disambiguation] <br /><br />
[https://wikis.swarthmore.edu/ling073/Kalmyk/Universal_Dependencies Universal Dependencies Syntax]<br />
<br />
=External Resources=<br />
== Computational Resources ==<br />
*online keyboard for Kalmyk [http://www.incks.com/en/kalmyk.html] <br /><br />
*another online keyboard for Kalymk, with two options - PC keyboard layout and the full alphabet layout [http://kalmyk.virtual-keyboards.com/]<br />
<br />
== Dictionaries ==<br />
*Kalmyk-Russian and Russian-Kalmyk online dictionary [http://www.multitran.ru/] <br /><br />
*Kalmyk-Russian dictionary, published in 1977; pdf format [http://altaica.ru/LIBRARY/muniev_1977.pdf] <br /><br />
*English-Kalmyk dictionary; student-created; pdf format [http://people.umass.edu/scable/LING404-SP16/Kalmyk/Kalmyk-Dictionary.pdf]<br /><br />
*English-Kalmyk dictionary; for some entries, contains description of word; licenses allow for free use ([https://glosbe.com/about]) [https://glosbe.com/en/xal/]<br /><br />
<br />
== Grammatical Descriptions ==<br />
*chapter describing overall grammar of Kalmyk, from the book ''The Mongolic Languages'' [https://www.dropbox.com/sh/0zd73b8lrky0f6z/AAC0ZCGQvqJk0HF4ofWCL2jza?dl=0]<br /><br />
*short description of very basic grammar [http://www.kalmykia.eu/2013/the-kalmyk-language-grammar-and-vocabulary/] <br /><br />
*English translation of "Cornelius Rahm's Kalmuk Grammar"; orthography and etymology [http://www.digizeitschriften.de/dms/img/?PID=PPN666048797_0013%7CLOG_0018&physid=PHYS_0109] <br /><br />
*archive of educational materials for preservation of Kalmyk, in Russian [http://biliq.ru/xalbook/]<br />
<br />
== Scientific Works ==<br />
*"Oirat Tobi: Intonational Structure of the Oirat Language"; includes some phonological analysis of Kalmyk [http://www.ling.hawaii.edu/graduate/Elena%20Indjieva%20Diss%20Draft.pdf] <br /><br />
<br />
== Corpora ==<br />
*wikipedia pages in Kalmyk [https://xal.wikipedia.org/wiki/%D0%A5%D0%B0%D0%BB%D1%8C%D0%BC%D0%B3] <br /><br />
*list of basic conversational phrases in Kalmyk [http://www.omniglot.com/language/phrases/kalmyk.htm] <br /><br />
*list of common words and phrases [https://en.wikivoyage.org/wiki/Kalmyk_phrasebook] <br /><br />
*the bible in Kalmyk; available in pdf and various ebook formats [https://ibtrussia.org/en/ebook?id=KAL] <br /><br />
*a memorial speech, with an English translation [https://www.fictionpress.com/s/3273887/1/%D0%A5%D0%B0%D0%BB%D1%8C%D0%BC%D0%B3-%D0%93%D0%B5%D0%BD%D0%BE%D1%86%D0%B8%D0%B4-%D0%94%D1%83%D1%80%D1%81%D1%85%D0%BB-%D2%AE%D0%B3] <br /><br />
*digital library of books in Kalmyk [http://biliq.ru/xalbook/library] <br />
<br />
[[Category:sp17_ResourceDocumentation]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4969
Kalmyk/Universal Dependencies
2017-04-29T00:14:12Z
<p>Slim1: /* Dependency Relations */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In my corpora, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "the people came eastward to the field of Shinar." Here, "field" has the nmod relation to "Shinar."<br />
<pre><br />
"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people <br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east<br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction<br />
"<Шинарин>"<br />
"Шинар" np gen @obj #4->6 !Shinar<br />
"<талвңд>"<br />
"талвң" n sg dat @nmod #5->4 !field <br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came<br />
</pre><br />
The following phrase translates to "we must mix (confuse) their language." Here, "language" has the nmod relation to "their."<br />
<pre><br />
"<Бидн>"<br />
"бидн" prn pers p1 pl inc nom @nsubj #6->9 !we<br />
"<теднә>"<br />
"тедн" prn pers p3 pl gen @amod #7->8 !they<br />
"<келинь>"<br />
"кел" n px3sp @nmod #8->7 !language<br />
"<хутхулхмн>"<br />
"хутхулхмн" v @root #9->0 !mix<br />
</pre><br />
<br />
===amod===<br />
amod is an adjective or an adjectival phrase that modifies a noun. This is a fairly straightforward relation that came up many times exactly as such in my corpora. <br /><br />
The following phrase translates to "Kalmyk nation" (Kalmykia). "Kalmyk" has the amod relation to "nation."<br />
<pre><br />
"<Хальмг>"<br />
"хальмг" adj @amod #9->10 !Kalmyk<br />
"<улс>"<br />
"улс" n sg nom @nsubj #10->13 !nation<br />
</pre><br />
The following phrase translates to "too few present (day) youths." "too few" and "present" both have the amod relation to "youths."<br />
<pre><br />
"<хор-хомс>"<br />
"хор-хомс" adj @amod #11->13 !too few<br />
"<одаһин>"<br />
"одаһин" adj @amod #12->13 !present<br />
"<баһчуд>"<br />
"баһчуд" n sg nom @nsubj #13->16 !youth<br />
</pre><br />
<br />
===det===<br />
det is a determiner head of a nominal. As expected, the words tagged as <det> had this head.<br />
The following phrase translates to "this day."<br />
<pre><br />
"<Эн>"<br />
"эн" det dem @det #1->2 !this<br />
"<өдр>"<br />
"өдр" n sg nom @nsubj #2->9 !day<br />
</pre><br />
The following phrase translates to "many people."<br />
<pre><br />
"<кесг>"<br />
"кесг" det @det #7->8 !many<br />
"<күн>"<br />
"күн" n sg nom @nsubj #8->10 !people<br />
</pre><br />
<br />
===obj===<br />
obj is the second most core argument of a verb. In the corpora, the nominal near the verb usually had this head, since Kalmyk is a SOV language.<br />
The following phrase translates to "this is called Operation Ulussy." "Operation" has the obj relation to the verb.<br />
<pre><br />
"<Эн>"<br />
"эн" det dem @nsubj #1->6 !this<br />
"<«>"<br />
"«" lquot @punct #2->3<br />
"<Операц>"<br />
"Операц" n sg nom @obj #3->6 !Operation<br />
"<Улусы>"<br />
"Улусы" np sg nom @nmod #4->3 !Ulussy<br />
"<»>"<br />
"»" rquot @punct #5->3<br />
"<нерлх>"<br />
"нерлх" v inf @root #6->0 !is called<br />
</pre><br />
<br />
The following phrase translates to "used tar instead of mortar" (the order of "tar" and "mortar" is reversed). <br />
<pre><br />
"<урсхлын>"<br />
"урсхлын" n sg gen @obj #5->8 !mortar<br />
"<ормд>"<br />
"ормд" prep @advmod #6->7 !instead of<br />
"<салм>"<br />
"салм" n sg nom @nsubj #7->8 !tar<br />
"<олзлв>"<br />
"олзл" v pret @root #8->0 !used<br />
</pre><br />
<br />
===acl===<br />
acl is used for finite and non-finite clauses that modify a nominal. In my corpora, acl were clauses that modified a noun.<br />
The following phrase translates to "they found the baby, who was lying in the manger." The core of the clause "lying in the manger" has the acl relation to "baby."<br />
<pre><br />
"<отльгт>"<br />
"отльгт" n sg nom @iobj #12->13 !manger<br />
"<кевтсн>"<br />
"кевт" v part past @acl #13->15 !lying<br />
"<теднə>"<br />
"теднə" prn pers p3 pl @nsubj #14->16 !they<br />
"<үринь>"<br />
"үр" n px3sp @obj #15->16 !baby<br />
"<үзцхəв>"<br />
"үзцх" v pret @root #16->9 !found<br />
</pre><br />
<br />
The following phrase translates to "God saw the city and tower that the people were building." The core of the clause "that the people were building" has the acl relation to "city."<br />
<pre><br />
"<әмтнә>"<br />
"әмтнә" n sg @obl #1->3 !people<br />
"<тосхҗах>"<br />
"тосхҗах" v fut @acl #2->3 !build<br />
"<балһс>"<br />
"балһс" n sg nom @conj #3->5 !city<br />
"<болн>"<br />
"болн" conjcoo @cc #4->5 !and<br />
"<бәәшң>"<br />
"бәәшң" n sg nom @csubj #5->6 !tower<br />
"<хәләхәр>"<br />
"хәләхәр" v @ccomp #6->10 !look<br />
"<Деед>"<br />
"Деед" adj @amod #7->8 !higher<br />
"<Эзн>"<br />
"эзн" n sg nom @nsubj #8->12 !master<br />
</pre><br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4968
Kalmyk/Universal Dependencies
2017-04-29T00:05:14Z
<p>Slim1: /* Dependency Relations */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In my corpora, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "the people came eastward to the field of Shinar." Here, "field" has the nmod relation to "Shinar."<br />
<pre><br />
"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people <br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east<br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction<br />
"<Шинарин>"<br />
"Шинар" np gen @obj #4->6 !Shinar<br />
"<талвңд>"<br />
"талвң" n sg dat @nmod #5->4 !field <br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came<br />
</pre><br />
The following phrase translates to "we must mix (confuse) their language." Here, "language" has the nmod relation to "their."<br />
<pre><br />
"<Бидн>"<br />
"бидн" prn pers p1 pl inc nom @nsubj #6->9 !we<br />
"<теднә>"<br />
"тедн" prn pers p3 pl gen @amod #7->8 !they<br />
"<келинь>"<br />
"кел" n px3sp @nmod #8->7 !language<br />
"<хутхулхмн>"<br />
"хутхулхмн" v @root #9->0 !mix<br />
</pre><br />
<br />
===amod===<br />
amod is an adjective or an adjectival phrase that modifies a noun. This is a fairly straightforward relation that came up many times exactly as such in my corpora. <br /><br />
The following phrase translates to "Kalmyk nation" (Kalmykia). "Kalmyk" has the amod relation to "nation."<br />
<pre><br />
"<Хальмг>"<br />
"хальмг" adj @amod #9->10 !Kalmyk<br />
"<улс>"<br />
"улс" n sg nom @nsubj #10->13 !nation<br />
</pre><br />
The following phrase translates to "too few present (day) youths." "too few" and "present" both have the amod relation to "youths."<br />
<pre><br />
"<хор-хомс>"<br />
"хор-хомс" adj @amod #11->13 !too few<br />
"<одаһин>"<br />
"одаһин" adj @amod #12->13 !present<br />
"<баһчуд>"<br />
"баһчуд" n sg nom @nsubj #13->16 !youth<br />
</pre><br />
<br />
===det===<br />
det is a determiner head of a nominal. As expected, the words tagged as <det> had this head.<br />
The following phrase translates to "this day."<br />
<pre><br />
"<Эн>"<br />
"эн" det dem @det #1->2 !this<br />
"<өдр>"<br />
"өдр" n sg nom @nsubj #2->9 !day<br />
</pre><br />
The following phrase translates to "many people."<br />
<pre><br />
"<кесг>"<br />
"кесг" det @det #7->8 !many<br />
"<күн>"<br />
"күн" n sg nom @nsubj #8->10 !people<br />
</pre><br />
<br />
===obj===<br />
obj is the second most core argument of a verb. In the corpora, the nominal near the verb usually had this head, since Kalmyk is a SOV language.<br />
The following phrase translates to "this is called Operation Ulussy." "Operation" has the obj relation to the verb.<br />
<pre><br />
"<Эн>"<br />
"эн" det dem @nsubj #1->6 !this<br />
"<«>"<br />
"«" lquot @punct #2->3<br />
"<Операц>"<br />
"Операц" n sg nom @obj #3->6 !Operation<br />
"<Улусы>"<br />
"Улусы" np sg nom @nmod #4->3 !Ulussy<br />
"<»>"<br />
"»" rquot @punct #5->3<br />
"<нерлх>"<br />
"нерлх" v inf @root #6->0 !is called<br />
</pre><br />
<br />
The following phrase translates to "used tar instead of mortar" (the order of "tar" and "mortar" is reversed). <br />
<pre><br />
"<урсхлын>"<br />
"урсхлын" n sg gen @obj #5->8 !mortar<br />
"<ормд>"<br />
"ормд" prep @advmod #6->7 !instead of<br />
"<салм>"<br />
"салм" n sg nom @nsubj #7->8 !tar<br />
"<олзлв>"<br />
"олзл" v pret @root #8->0 !used<br />
</pre><br />
<br />
===acl===<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4967
Kalmyk/Universal Dependencies
2017-04-28T23:57:31Z
<p>Slim1: /* Dependency Relations */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In my corpora, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "the people came eastward to the field of Shinar." Here, "field" has the nmod relation with regards to "Shinar."<br />
<pre><br />
"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people <br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east<br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction<br />
"<Шинарин>"<br />
"Шинар" np gen @obj #4->6 !Shinar<br />
"<талвңд>"<br />
"талвң" n sg dat @nmod #5->4 !field <br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came<br />
</pre><br />
The following phrase translates to "we must mix (confuse) their language." Here, "language" has the nmod relation with regards to "their."<br />
<pre><br />
"<Бидн>"<br />
"бидн" prn pers p1 pl inc nom @nsubj #6->9 !we<br />
"<теднә>"<br />
"тедн" prn pers p3 pl gen @amod #7->8 !they<br />
"<келинь>"<br />
"кел" n px3sp @nmod #8->7 !language<br />
"<хутхулхмн>"<br />
"хутхулхмн" v @root #9->0 !mix<br />
</pre><br />
<br />
===amod===<br />
amod is an adjective or an adjectival phrase that modifies a noun. This is a fairly straightforward relation that came up many times exactly as such in my corpora. <br /><br />
The following phrase translates to "Kalmyk nation" (Kalmykia). "Kalmyk" has the amod relation with regards to "nation."<br />
<pre><br />
"<Хальмг>"<br />
"хальмг" adj @amod #9->10 !Kalmyk<br />
"<улс>"<br />
"улс" n sg nom @nsubj #10->13 !nation<br />
</pre><br />
The following phrase translates to "too few present (day) youths." "too few" and "present" both have the amod relation with regards to "youths."<br />
<pre><br />
"<хор-хомс>"<br />
"хор-хомс" adj @amod #11->13 !too few<br />
"<одаһин>"<br />
"одаһин" adj @amod #12->13 !present<br />
"<баһчуд>"<br />
"баһчуд" n sg nom @nsubj #13->16 !youth<br />
</pre><br />
<br />
===det===<br />
det is a determiner head of a nominal. As expected, the words tagged as <det> had this head.<br />
The following phrase translates to "this day."<br />
<pre><br />
"<Эн>"<br />
"эн" det dem @det #1->2 !this<br />
"<өдр>"<br />
"өдр" n sg nom @nsubj #2->9 !day<br />
</pre><br />
The following phrase translates to "many people."<br />
<pre><br />
"<кесг>"<br />
"кесг" det @det #7->8 !many<br />
"<күн>"<br />
"күн" n sg nom @nsubj #8->10 !people<br />
</pre><br />
<br />
===obj===<br />
<br />
===acl===<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4966
Kalmyk/Universal Dependencies
2017-04-28T22:43:06Z
<p>Slim1: /* Dependency Relations */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In my corpora, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "the people came eastward to the field of Shinar." Here, "field" has the nmod relation with regards to "Shinar."<br />
<pre><br />
"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people <br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east<br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction<br />
"<Шинарин>"<br />
"Шинар" np gen @obj #4->6 !Shinar<br />
"<талвңд>"<br />
"талвң" n sg dat @nmod #5->4 !field <br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came<br />
</pre><br />
The following phrase translates to "we must mix (confuse) their language." Here, "language" has the nmod relation with regards to "their."<br />
<pre><br />
"<Бидн>"<br />
"бидн" prn pers p1 pl inc nom @nsubj #6->9 !we<br />
"<теднә>"<br />
"тедн" prn pers p3 pl gen @amod #7->8 !they<br />
"<келинь>"<br />
"кел" n px3sp @nmod #8->7 !language<br />
"<хутхулхмн>"<br />
"хутхулхмн" v @root #9->0 !mix<br />
</pre><br />
<br />
===amod===<br />
amod is an adjective or an adjectival phrase that modifies a noun. This is a fairly straightforward relation that came up many times exactly as such in my corpora. <br /><br />
The following phrase translates to "Kalmyk nation" (Kalmykia). "Kalmyk" has the amod relation with regards to "nation."<br />
<pre><br />
"<Хальмг>"<br />
"хальмг" adj @amod #9->10 !Kalmyk<br />
"<улс>"<br />
"улс" n sg nom @nsubj #10->13 !nation<br />
</pre><br />
The following phrase translates to "too few present (day) youths." "too few" and "present" both have the amod relation with regards to "youths."<br />
<pre><br />
"<хор-хомс>"<br />
"хор-хомс" adj @amod #11->13 !too few<br />
"<одаһин>"<br />
"одаһин" adj @amod #12->13 !present<br />
"<баһчуд>"<br />
"баһчуд" n sg nom @nsubj #13->16 !youth<br />
</pre><br />
<br />
===det===<br />
<br />
===obj===<br />
<br />
===acl===<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4965
Kalmyk/Universal Dependencies
2017-04-28T22:26:46Z
<p>Slim1: /* nmod */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In my corpora, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "the people came eastward to the field of Shinar." Here, "field" has the nmod relation with regards to "Shinar."<br />
<pre><br />
"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people <br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east<br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction<br />
"<Шинарин>"<br />
"Шинар" np gen @obj #4->6 !Shinar<br />
"<талвңд>"<br />
"талвң" n sg dat @nmod #5->4 !field <br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came<br />
</pre><br />
The following phrase translates to "we must mix (confuse) their language." Here, "language" has the nmod relation with regards to "their."<br />
<pre><br />
"<Бидн>"<br />
"бидн" prn pers p1 pl inc nom @nsubj #6->9 !we<br />
"<теднә>"<br />
"тедн" prn pers p3 pl gen @amod #7->8 !they<br />
"<келинь>"<br />
"кел" n px3sp @nmod #8->7 !language<br />
"<хутхулхмн>"<br />
"хутхулхмн" v @root #9->0 !mix<br />
</pre><br />
<br />
===amod===<br />
<br />
===det===<br />
<br />
===obj===<br />
<br />
===acl===<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4964
Kalmyk/Universal Dependencies
2017-04-28T22:23:31Z
<p>Slim1: /* nmod */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In my corpora, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "the people came eastward to the plain of Shinar." <br />
<pre><br />
"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people <br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east<br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction<br />
"<Шинарин>"<br />
"Шинар" np gen @obj #4->6 !Shinar<br />
"<талвңд>"<br />
"талвң" n sg dat @nmod #5->4 !field <br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came <br /><br />
</pre><br />
<br />
===amod===<br />
<br />
===det===<br />
<br />
===obj===<br />
<br />
===acl===<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4963
Kalmyk/Universal Dependencies
2017-04-28T22:21:27Z
<p>Slim1: /* nmod */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In my corpora, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "the people came eastward to the plain of Shinar." <br />
<code><br />
"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people <br /><br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east <br /><br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction <br /><br />
"<Шинарин>"<br />
"Шинар" np gen @nmod #4->5 !Shinar's <br /><br />
"<талвңд>"<br />
"талвң" n sg dat @obj #5->6 !field <br /><br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came <br /><br />
</code><br />
<br />
===amod===<br />
<br />
===det===<br />
<br />
===obj===<br />
<br />
===acl===<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4962
Kalmyk/Universal Dependencies
2017-04-28T22:18:45Z
<p>Slim1: /* Dependency Relations */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
===nmod===<br />
nmod is used for nominal dependents of another noun. In Kalmyk, this relation was frequently used for the genitive construction. <br /><br />
The following phrase translates to "<br />
<code>"<Әмтн>"<br />
"әмтн" n sg nom @nsubj #1->6 !people<br />
"<дорд>"<br />
"дорд" n sg nom @appos #2->3 !east<br />
"<үзгәс>"<br />
"үзс" n sg abl @amod #3->6 !direction<br />
"<Шинарин>"<br />
"Шинар" np gen @nmod #4->5 !Shinar's<br />
"<талвңд>"<br />
"талвң" n sg dat @obj #5->6 !field<br />
"<ирәд>"<br />
"ирәд" v pret @root #6->0 !came<br />
</code><br />
<br />
<br />
===amod===<br />
<br />
===det===<br />
<br />
===obj===<br />
<br />
===acl===<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4959
Kalmyk/Universal Dependencies
2017-04-28T03:26:08Z
<p>Slim1: </p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! xal.withmorph.udpipe<br />
! xal.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! xal.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! xal.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4958
Kalmyk and Judeo-Tat
2017-04-28T03:11:21Z
<p>Slim1: /* Transducer Evaluation */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision: 100.00%<br />
*recall: 47.92%<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision: 99.32886%<br />
*recall: 68.51852%<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus: 2416<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus: 16/104<br />
*trimmed coverage over <code>longer</code> corpus: 24.00%<br />
*trimmed coverage over <code>large</code> corpus: 23.22%<br />
*number of tokens in <code>longer</code> corpus: 110<br />
*number of tokens in <code>large</code> corpus: 2461<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus: 24/101<br />
*trimmed coverage over <code>longer</code> corpus: 29.77%<br />
*trimmed coverage over <code>large</code> corpus: 21.12%<br />
*number of tokens in <code>longer</code> corpus: 101<br />
*number of tokens in <code>large</code> corpus: 4721<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4952
Kalmyk/Universal Dependencies
2017-04-28T01:40:59Z
<p>Slim1: /* Evaluation */</p>
<hr />
<div>=Evaluation=<br />
{| class="wikitable"<br />
|+ UD Parser Evaluation<br />
|-<br />
! <br />
! lin.withmorph.udpipe<br />
! lin.nomorph.udpipe<br />
! Number of forms<br />
! Number of sentences<br />
|-<br />
! lin.annotated.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|-<br />
! lin.annotated2.ud.conllu<br />
| UAS: TODO, LAS: TODO<br />
| UAS: TODO, LAS: TODO<br />
| TODO<br />
| TODO<br />
|}<br />
<br />
=Dependency Relations=<br />
<br />
<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk/Universal_Dependencies&diff=4951
Kalmyk/Universal Dependencies
2017-04-28T01:39:18Z
<p>Slim1: Created page with "=Evaluation= =Dependency Relations= Category:sp17_UD Category:Kalmyk"</p>
<hr />
<div>=Evaluation=<br />
<br />
<br />
<br />
=Dependency Relations=<br />
<br />
<br />
<br />
<br />
[[Category:sp17_UD]]<br />
[[Category:Kalmyk]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4774
Kalmyk and Judeo-Tat
2017-04-23T22:47:40Z
<p>Slim1: /* Transducer Evaluation */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision: 100.00%<br />
*recall: 47.92%<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus: 2416<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus: 16/104<br />
*trimmed coverage over <code>longer</code> corpus: 24.00%<br />
*trimmed coverage over <code>large</code> corpus: 23.22%<br />
*number of tokens in <code>longer</code> corpus: 110<br />
*number of tokens in <code>large</code> corpus: 2461<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus: 24/101<br />
*trimmed coverage over <code>longer</code> corpus: 29.77%<br />
*trimmed coverage over <code>large</code> corpus: 21.12%<br />
*number of tokens in <code>longer</code> corpus: 101<br />
*number of tokens in <code>large</code> corpus: 4721<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4763
Kalmyk and Judeo-Tat
2017-04-22T20:18:21Z
<p>Slim1: /* jdt->xal */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus: 2416<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus: 16/104<br />
*trimmed coverage over <code>longer</code> corpus: 24.00%<br />
*trimmed coverage over <code>large</code> corpus: 23.22%<br />
*number of tokens in <code>longer</code> corpus: 110<br />
*number of tokens in <code>large</code> corpus: 2461<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus: 24/101<br />
*trimmed coverage over <code>longer</code> corpus: 29.77%<br />
*trimmed coverage over <code>large</code> corpus: 21.12%<br />
*number of tokens in <code>longer</code> corpus: 101<br />
*number of tokens in <code>large</code> corpus: 4721<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4762
Kalmyk and Judeo-Tat
2017-04-22T20:17:06Z
<p>Slim1: /* MT Evaluation */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus: 2416<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*trimmed coverage over <code>longer</code> corpus: 24.00%<br />
*trimmed coverage over <code>large</code> corpus: 23.22%<br />
*number of tokens in <code>longer</code> corpus: 110<br />
*number of tokens in <code>large</code> corpus: 2461<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus: 24/101<br />
*trimmed coverage over <code>longer</code> corpus: 29.77%<br />
*trimmed coverage over <code>large</code> corpus: 21.12%<br />
*number of tokens in <code>longer</code> corpus: 101<br />
*number of tokens in <code>large</code> corpus: 4721<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4761
Kalmyk and Judeo-Tat
2017-04-22T19:53:46Z
<p>Slim1: /* MT Evaluation */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus: 2416<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*trimmed coverage over <code>longer</code> corpus: 24.00%<br />
*trimmed coverage over <code>large</code> corpus: 23.22%<br />
*number of tokens in <code>longer</code> corpus: 110<br />
*number of tokens in <code>large</code> corpus: 2461<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*trimmed coverage over <code>longer</code> corpus: 29.77%<br />
*trimmed coverage over <code>large</code> corpus: 21.12%<br />
*number of tokens in <code>longer</code> corpus: 101<br />
*number of tokens in <code>large</code> corpus: 4721<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4760
Kalmyk and Judeo-Tat
2017-04-22T19:35:23Z
<p>Slim1: /* jdt->xal */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus: 2416<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*trimmed coverage over <code>longer</code> corpus:<br />
*trimmed coverage over <code>large</code> corpus:<br />
*number of tokens in <code>longer</code> corpus: 110<br />
*number of tokens in <code>large</code> corpus: 2461<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*trimmed coverage over <code>longer</code> corpus:<br />
*trimmed coverage over <code>large</code> corpus:<br />
*number of tokens in <code>longer</code> corpus: 101<br />
*number of tokens in <code>large</code> corpus: 4721<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4759
Kalmyk and Judeo-Tat
2017-04-22T19:33:07Z
<p>Slim1: /* xal->jdt */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus: 2416<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*trimmed coverage over <code>longer</code> corpus:<br />
*trimmed coverage over <code>large</code> corpus:<br />
*number of tokens in <code>longer</code> corpus: 101<br />
*number of tokens in <code>large</code> corpus: 4721<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4758
Kalmyk and Judeo-Tat
2017-04-22T18:48:16Z
<p>Slim1: /* Transducer Evaluation */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus: 2416<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4757
Kalmyk and Judeo-Tat
2017-04-22T18:47:55Z
<p>Slim1: /* Transducer Evaluation */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 4709<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus:<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4756
Kalmyk and Judeo-Tat
2017-04-22T18:40:13Z
<p>Slim1: /* Transducer Evaluation */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 58060<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 39.57%<br />
*number of words in <code>large</code> corpus:<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1
http://wikis.swarthmore.edu/ling073/index.php?title=Kalmyk_and_Judeo-Tat&diff=4755
Kalmyk and Judeo-Tat
2017-04-22T18:22:54Z
<p>Slim1: /* jdt->xal */</p>
<hr />
<div>Resources for machine translation between [https://wikis.swarthmore.edu/ling073/Kalmyk Kalmyk] and [https://wikis.swarthmore.edu/ling073/Judeo-Tat Judeo-Tat]<br />
=Contrastive Grammar=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Contrastive_Grammar Contrastive Grammar Page]<br />
<br />
=Parallel Corpus=<br />
[https://github.swarthmore.edu/slim1/ling073-xal-corpus Parallel Corpus on GitHub]<br />
<br />
=Lexical Selection=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Lexical_Selection Lexical Selection page] <br /><br />
==jdt → xal==<br />
* (jdt) бирор → (xal) ах (older); дүү (younger) "brother"<br />
* (jdt) хэьер → (xal) эгч (older); дү күүкн (younger) "sister"<br />
<br />
==xal → jdt==<br />
* (xal) бәәх (live; stand; stop) → (jdt) дуланмиш (live); поис (stand) <br />
* (xal) зүркн (heart) → (jdt) гъирйет (courage); дуьл (heart)<br />
<br />
= jdt → xal evaluation = <br />
WER and PER are both 100% even though it should be translating properly.<br />
<br />
<code>jdt.sentences.txt</code> has 102 words, of which 88 are unknown.<br />
<code>xal.sentences.txt</code> has 99 words, of which 79 are unknown.<br />
<br />
=Structural Transfer=<br />
[https://wikis.swarthmore.edu/ling073/Kalmyk_and_Judeo-Tat/Structural_transfer Structural Transfer page]<br />
<br />
= Final Evaluation =<br />
We have added 200 stems.<br />
<pre><br />
<e><p><l>пул<s n="n"/></l><r>мөңгн<s n="n"/></r></p></e><br />
<e><p><l>пэхьни<s n="adj"/></l><r>далд<s n="adj"/></r></p></e><br />
<e><p><l>чапар<s n="n"/></l><r>хаша<s n="n"/></r></p></e><br />
<e><p><l>сохъ<s n="adj"/></l><r>бүкл<s n="adj"/></r></p></e><br />
<e><p><l>сэгӀэт<s n="n"/></l><r>цаг<s n="n"/></r></p></e><br />
<e><p><l>шогъгъо<s n="v"/></l><r>доңhдулх<s n="v"/></r></p></e><br />
<e><p><l>хэрегуш<s n="n"/></l><r>кирулә<s n="n"/></r></p></e><br />
<e><p><l>овчи<s n="n"/></l><r>аңhуч<s n="n"/></r></p></e><br />
<e><p><l>вирих<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>менэхьэфи<s n="n"/></l><r>магтал<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>сег<s n="n"/></r></p></e><br />
<e><p><l>гоф<s n="n"/></l><r>үг<s n="n"/></r></p></e><br />
<e><p><l>верф<s n="n"/></l><r>цасн<s n="n"/></r></p></e><br />
<e><p><l>хори<s n="n"/></l><r>hазр<s n="n"/></r></p></e><br />
<e><p><l>руш<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>бигъ<s n="n"/></l><r>сахл<s n="n"/></r></p></e><br />
<e><p><l>огъо<s n="n"/></l><r>эзн<s n="n"/></r></p></e><br />
<e><p><l>шори<s n="n"/></l><r>байр<s n="n"/></r></p></e><br />
<e><p><l>ники<s n="n"/></l><r>буйн<s n="n"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>нертә<s n="adj"/></r></p></e><br />
<e><p><l>мэгӀлуьм<s n="adj"/></l><r>алдр<s n="adj"/></r></p></e><br />
<e><p><l>гӀуьом<s n="n"/></l><r>делкә<s n="n"/></r></p></e><br />
<e><p><l>те<s n="pr"/></l><r>күртл<s n="pr"/></r></p></e><br />
<e><p><l>имогьо<s n="adv"/></l><r>ода<s n="adv"/></r></p></e><br />
<e><p><l>вераси<s n="v"/></l><r>даминь авх<s n="v"/></r></p></e><br />
<e><p><l>вози<s n="v"/></l><r>играть<s n="v"/></r></p></e><br />
<e><p><l>пушо<s n="pr"/></l><r>өмн<s n="pr"/></r></p></e><br />
<e><p><l>гqэил<s n="n"/></l><r>күүкд<s n="n"/></r></p></e><br />
<e><p><l>пушоте<s n="adj"/></l><r>эрт<s n="adj"/></r></p></e><br />
<e><p><l>хьэйф<s n="n"/></l><r>өр өвдлhн<s n="n"/></r></p></e><br />
<e><p><l>гушд<s n="n"/></l><r>мясо<s n="n"/></r></p></e><br />
<e><p><l>эрйьси<s n="n"/></l><r>хүрм<s n="n"/></r></p></e><br />
<e><p><l>гов<s n="n"/></l><r>үкр<s n="n"/></r></p></e><br />
<e><p><l>овчийети<s n="v"/></l><r>аңhучлх<s n="v"/></r></p></e><br />
<e><p><l>више<s n="n"/></l><r>ө-модн<s n="n"/></r></p></e><br />
<e><p><l>дере<s n="n"/></l><r>огчм<s n="n"/></r></p></e><br />
<e><p><l>фурмун<s n="v"/></l><r>меклх<s n="v"/></r></p></e><br />
<e><p><l>духдер<s n="n"/></l><r>күүкн<s n="n"/></r></p></e><br />
<e><p><l>шев<s n="n"/></l><r>сө<s n="n"/></r></p></e><br />
<e><p><l>шйьвер<s n="n"/></l><r>залу<s n="n"/></r></p></e><br />
<e><p><l>тэхьно<s n="adj"/></l><r>hанцарн<s n="adj"/></r></p></e><br />
<e><p><l>гьемише<s n="n"/></l><r>даң<s n="n"/></r></p></e><br />
<e><p><l>бирэхьне<s n="n"/></l><r>hолдhр<s n="n"/></r></p></e><br />
<e><p><l>гъэевош<s n="n"/></l><r>зарц күүкн<s n="n"/></r></p></e><br />
<e><p><l>бике<s n="n"/></l><r>керл<s n="n"/></r></p></e><br />
<e><p><l>чуь<s n="prn"/></l><r>что<s n="prn"/></r></p></e><br />
<e><p><l>дерму<s n="n"/></l><r>медицин<s n="n"/></r></p></e><br />
<e><p><l>хис<s n="v"/></l><r>унтх<s n="v"/></r></p></e><br />
<e><p><l>суьфре<s n="n"/></l><r>таблиц<s n="n"/></r></p></e><br />
<e><p><l>пинеши<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>хуб<s n="adv"/></l><r>сәәнәр<s n="adv"/></r></p></e><br />
<e><p><l>варайун<s n="v"/></l><r>зокъялх<s n="v"/></r></p></e><br />
<e><p><l>дивор<s n="n"/></l><r>эрс<s n="n"/></r></p></e><br />
<e><p><l>йон<s n="n"/></l><r>талк<s n="n"/></r></p></e><br />
<e><p><l>сифет<s n="n"/></l><r>амн-хамр<s n="n"/></r></p></e><br />
<e><p><l>сипре<s n="n"/></l><r>цаhан<s n="n"/></r></p></e><br />
<e><p><l>ки<s n="prn"/></l><r>кен<s n="prn"/></r></p></e><br />
<e><p><l>бур<s n="v"/></l><r>алх<s n="v"/></r></p></e><br />
<e><p><l>пуч<s n="n"/></l><r>әрлhәх<s n="n"/></r></p></e><br />
<e><p><l>хуг<s n="n"/></l><r>hаха<s n="n"/></r></p></e><br />
<e><p><l>хун<s n="n"/></l><r>цусн<s n="n"/></r></p></e><br />
<e><p><l>халиф<s n="n"/></l><r>халиф<s n="n"/></r></p></e><br />
<e><p><l>гьеле<s n="adv"/></l><r>бас<s n="adv"/></r></p></e><br />
<e><p><l>ижире<s n="det"/></l><r>иим<s n="det"/></r></p></e><br />
<e><p><l>рач<s n="adj"/></l><r>айта<s n="adj"/></r></p></e><br />
<e><p><l>вемаси<s n="adj"/></l><r>бадярта<s n="adj"/></r></p></e><br />
<e><p><l>шуьгъэм<s n="n"/></l><r>гесн<s n="n"/></r></p></e><br />
<e><p><l>озгъун<s n="adj"/></l><r>әәмшгтә<s n="adj"/></r></p></e><br />
<e><p><l>жоллод<s n="n"/></l><r>алач<s n="n"/></r></p></e><br />
<e><p><l>шимшил<s n="n"/></l><r>селм<s n="n"/></r></p></e><br />
<e><p><l>офдо<s n="v"/></l><r>буурах<s n="v"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>дәкн<s n="n"/></r></p></e><br />
<e><p><l>гене<s n="n"/></l><r>бас<s n="n"/></r></p></e><br />
<e><p><l>шовгӀо<s n="n"/></l><r>андhар<s n="n"/></r></p></e><br />
<e><p><l>хур<s n="v"/></l><r>идх<s n="v"/></r></p></e><br />
<e><p><l>бэгъ<s n="n"/></l><r>меклә<s n="n"/></r></p></e><br />
<e><p><l>гӀэсел<s n="n"/></l><r>бал<s n="n"/></r></p></e><br />
<e><p><l>мигӀид<s n="n"/></l><r>сән өдрллhн<s n="n"/></r></p></e><br />
<e><p><l>карасди<s n="n"/></l><r>зевсг<s n="n"/></r></p></e><br />
<e><p><l>музике<s n="n"/></l><r>көгҗм<s n="n"/></r></p></e><br />
<e><p><l>мэгӀни<s n="n"/></l><r>дун<s n="n"/></r></p></e><br />
<e><p><l>куьф<s n="n"/></l><r>хавтх<s n="n"/></r></p></e><br />
<e><p><l>корде<s n="n"/></l><r>утх<s n="n"/></r></p></e><br />
<e><p><l>мейит<s n="n"/></l><r>үксн бий<s n="n"/></r></p></e><br />
<e><p><l>зури-зури<s n="adv"/></l><r>түргн<s n="adv"/></r></p></e><br />
<e><p><l>имид<s n="v"/></l><r>ицх<s n="v"/></r></p></e><br />
<e><p><l>гирисде<s n="v"/></l><r>уульх<s n="v"/></r></p></e><br />
<e><p><l>рай<s n="n"/></l><r>хурц ухан<s n="n"/></r></p></e><br />
<e><p><l>овур<s n="v"/></l><r>асрх<s n="v"/></r></p></e><br />
<e><p><l>пой<s n="n"/></l><r>Көл<s n="n"/></r></p></e><br />
<e><p><l>гӀэребе<s n="n"/></l><r>тергн<s n="n"/></r></p></e><br />
<e><p><l>видовус<s n="v"/></l><r>гүүх<s n="v"/></r></p></e><br />
<e><p><l>хэрмон<s n="v"/></l><r>цокх<s n="v"/></r></p></e><br />
<e><p><l>чул<s n="n"/></l><r>девән<s n="n"/></r></p></e><br />
<e><p><l>дворник<s n="n"/></l><r>үүдч<s n="n"/></r></p></e><br />
<e><p><l>куче<s n="n"/></l><r>уульнцин<s n="n"/></r></p></e><br />
<e><p><l>партал<s n="n"/></l><r>хувцн<s n="n"/></r></p></e><br />
<e><p><l>шуш<s n="v"/></l><r>уhах<s n="v"/></r></p></e><br />
<e><p><l>кук<s n="n"/></l><r>көвүн<s n="n"/></r></p></e><br />
<e><p><l>завод<s n="n"/></l><r>эд<s n="n"/></r></p></e><br />
<e><p><l>шише<s n="n"/></l><r>ширк<s n="n"/></r></p></e><br />
<e><p><l>инженер<s n="n"/></l><r>Инженер<s n="n"/></r></p></e><br />
<e><p><l>тсех<s n="n"/></l><r>мастерской<s n="n"/></r></p></e><br />
<e><p><l>рэхьбер<s n="v"/></l><r>зүглүлх<s n="v"/></r></p></e><br />
<e><p><l>пенжере<s n="n"/></l><r>терз<s n="n"/></r></p></e><br />
<e><p><l>аслан<s n="n"/></l><r>арслң<s n="n"/></r></p></e><br />
<e><p><l>йобу<s n="adv"/></l><r>му<s n="adv"/></r></p></e><br />
<e><p><l>хов<s n="n"/></l><r>зүүдн<s n="n"/></r></p></e><br />
<e><p><l>туьлки<s n="n"/></l><r>арат<s n="n"/></r></p></e><br />
<e><p><l>бегьем<s n="v"/></l><r>күрх<s n="v"/></r></p></e><br />
<e><p><l>чай<s n="n"/></l><r>цә<s n="n"/></r></p></e><br />
<e><p><l>чуп<s n="n"/></l><r>кружк<s n="n"/></r></p></e><br />
<e><p><l>хьуьзуьм<s n="n"/></l><r>модн<s n="n"/></r></p></e><br />
<e><p><l>герми<s n="n"/></l><r>дулаhар<s n="n"/></r></p></e><br />
<e><p><l>гӀорд<s n="n"/></l><r>hуйр<s n="n"/></r></p></e><br />
<e><p><l>билесумол<s n="n"/></l><r>будгта<s n="n"/></r></p></e><br />
<e><p><l>йесийовчи<s n="n"/></l><r>теермч<s n="n"/></r></p></e><br />
<e><p><l>гӀэдот<s n="n"/></l><r>авъяс<s n="n"/></r></p></e><br />
<e><p><l>олхогьини<s n="n"/></l><r>бах<s n="n"/></r></p></e><br />
<e><p><l>темиз<s n="adj"/></l><r>арчлhн<s n="adj"/></r></p></e><br />
<e><p><l>тем<s n="n"/></l><r>бий<s n="n"/></r></p></e><br />
<e><p><l>кусе<s n="n"/></l><r>сахлго<s n="n"/></r></p></e><br />
<e><p><l>дерйо<s n="n"/></l><r>теңгс<s n="n"/></r></p></e><br />
<e><p><l>шохьонгум<s n="n"/></l><r>асхн<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>йох<s n="n"/></r></p></e><br />
<e><p><l>васал<s n="n"/></l><r>хавр<s n="n"/></r></p></e><br />
<e><p><l>верас<s n="v"/></l><r>гүүцәх<s n="v"/></r></p></e><br />
<e><p><l>тум<s n="n"/></l><r>эрднь-иш<s n="n"/></r></p></e><br />
<e><p><l>женй<s n="n"/></l><r>бәрлдәнә<s n="n"/></r></p></e><br />
<e><p><l>хьэрекет<s n="n"/></l><r>меңслhн<s n="n"/></r></p></e><br />
<e><p><l>гӀэрмудор<s n="n"/></l><r>кедмнә модн<s n="n"/></r></p></e><br />
<e><p><l>гисне<s n="adj"/></l><r>харhнсн<s n="adj"/></r></p></e><br />
<e><p><l>нун<s n="n"/></l><r>өдмг<s n="n"/></r></p></e><br />
<e><p><l>вохур<s n="v"/></l><r>харhлдх<s n="v"/></r></p></e><br />
<e><p><l>пуьрси<s n="v"/></l><r>сурх<s n="v"/></r></p></e><br />
<e><p><l>чекмечи<s n="n"/></l><r>hосч<s n="n"/></r></p></e><br />
<e><p><l>сер<s n="n"/></l><r>толhа<s n="n"/></r></p></e><br />
<e><p><l>нэхичи<s n="n"/></l><r>малч<s n="n"/></r></p></e><br />
<e><p><l>гӀэйб<s n="n"/></l><r>ичр<s n="n"/></r></p></e><br />
<e><p><l>гомиш<s n="n"/></l><r>маҗ<s n="n"/></r></p></e><br />
<e><p><l>жэгӀмет<s n="n"/></l><r>община<s n="n"/></r></p></e><br />
<e><p><l>ошне<s n="n"/></l><r>иньглсн<s n="n"/></r></p></e><br />
<e><p><l>игид<s n="n"/></l><r>баатр<s n="n"/></r></p></e><br />
<e><p><l>герден<s n="n"/></l><r>күзүн<s n="n"/></r></p></e><br />
<e><p><l>согъ<s n="adj"/></l><r>эрүл<s n="adj"/></r></p></e><br />
<e><p><l>гьэгъ<s n="n"/></l><r>җалв<s n="n"/></r></p></e><br />
<e><p><l>согъсаламат<s n="n"/></l><r>амулң<s n="n"/></r></p></e><br />
<e><p><l>суьрх<s n="n"/></l><r>алтн<s n="n"/></r></p></e><br />
<e><p><l>егенер<s n="conj"/></l><r>болхла<s n="conj"/></r></p></e><br />
<e><p><l>гьер<s n="det"/></l><r>болhн<s n="det"/></r></p></e><br />
<e><p><l>фенд<s n="n"/></l><r>хар<s n="n"/></r></p></e><br />
<e><p><l>гӀэмел<s n="n"/></l><r>ухалҗ<s n="n"/></r></p></e><br />
<e><p><l>жуьргӀэт<s n="n"/></l><r>сөдрмг<s n="n"/></r></p></e><br />
<e><p><l>хэйли<s n="adv"/></l><r>йир<s n="adv"/></r></p></e><br />
<e><p><l>торик<s n="adj"/></l><r>хар<s n="adj"/></r></p></e><br />
<e><p><l>мар<s n="n"/></l><r>моhа<s n="n"/></r></p></e><br />
<e><p><l>руьгъэ<s n="n"/></l><r>өрмин<s n="n"/></r></p></e><br />
<e><p><l>оммо<s n="pr"/></l><r>талдан<s n="pr"/></r></p></e><br />
<e><p><l>фэхьлеи<s n="n"/></l><r>аҗлх<s n="n"/></r></p></e><br />
<e><p><l>зэхьмет<s n="n"/></l><r>өвдкүр<s n="n"/></r></p></e><br />
<e><p><l>жофо<s n="n"/></l><r>аҗл<s n="n"/></r></p></e><br />
<e><p><l>веровун<s n="v"/></l><r>серлх<s n="v"/></r></p></e><br />
<e><p><l>гъимет<s n="n"/></l><r>үн<s n="n"/></r></p></e><br />
<e><p><l>огол<s n="v"/></l><r>нерәдсн<s n="v"/></r></p></e><br />
<e><p><l>шувун<s n="v"/></l><r>саҗх<s n="v"/></r></p></e><br />
<e><p><l>келемерди<s n="adj"/></l><r>көгшн<s n="adj"/></r></p></e><br />
<e><p><l>бала<s n="n"/></l><r>өндр<s n="n"/></r></p></e><br />
<e><p><l>типик<s n="adj"/></l><r>боса<s n="adj"/></r></p></e><br />
<e><p><l>жогьил<s n="n"/></l><r>идр<s n="n"/></r></p></e><br />
<e><p><l>ведилуь<s n="n"/></l><r>алдр<s n="n"/></r></p></e><br />
<e><p><l>гъуморбаз<s n="n"/></l><r>наадач<s n="n"/></r></p></e><br />
<e><p><l>жон<s n="adj"/></l><r>әрүхн<s n="adj"/></r></p></e><br />
<e><p><l>шилоли<s n="n"/></l><r>харцх<s n="n"/></r></p></e><br />
<e><p><l>мужак<s n="n"/></l><r>бөкүн<s n="n"/></r></p></e><br />
<e><p><l>жэгӀчи<s n="n"/></l><r>заhсч<s n="n"/></r></p></e><br />
<e><p><l>тур<s n="n"/></l><r>шүүгүл<s n="n"/></r></p></e><br />
<e><p><l>вешсем<s n="v"/></l><r>өргх<s n="v"/></r></p></e><br />
<e><p><l>кекуьл<s n="n"/></l><r>шил<s n="n"/></r></p></e><br />
<e><p><l>тимор<s n="n"/></l><r>цоклт<s n="n"/></r></p></e><br />
<e><p><l>гези<s n="n"/></l><r>зуулhн<s n="n"/></r></p></e><br />
<e><p><l>руьгъэни<s n="adj"/></l><r>әрмгтә<s n="adj"/></r></p></e><br />
<e><p><l>рэхьо<s n="n"/></l><r>сулдхл<s n="n"/></r></p></e><br />
<e><p><l>пурр<s n="n"/></l><r>бүкл<s n="n"/></r></p></e><br />
<e><p><l>дохьори<s n="n"/></l><r>чолун<s n="n"/></r></p></e><br />
<e><p><l>симов<s n="v"/></l><r>өөмх<s n="v"/></r></p></e><br />
<e><p><l>батмиш<s n="v"/></l><r>чивх<s n="v"/></r></p></e><br />
<e><p><l>чуьшме<s n="n"/></l><r>нарн<s n="n"/></r></p></e><br />
<e><p><l>гъуши<s n="n"/></l><r>церг<s n="n"/></r></p></e><br />
<e><p><l>мешат<s n="n"/></l><r>саалтг<s n="n"/></r></p></e><br />
<e><p><l>хэн<s n="v"/></l><r>инәдн<s n="v"/></r></p></e><br />
<e><p><l>вини<s n="n"/></l><r>хамр<s n="n"/></r></p></e><br />
<e><p><l>сес<s n="n"/></l><r>ә<s n="n"/></r></p></e><br />
<e><p><l>дирногъ<s n="n"/></l><r>турун<s n="n"/></r></p></e><br />
<e><p><l>тевер<s n="n"/></l><r>сүк<s n="n"/></r></p></e><br />
<e><p><l>номус<s n="n"/></l><r>күнд<s n="n"/></r></p></e><br />
<e><p><l>расди<s n="n"/></l><r>үнн<s n="n"/></r></p></e><br />
<e><p><l>дивэгӀ<s n="n"/></l><r>йиртмҗ<s n="n"/></r></p></e><br />
<e><p><l>бебе<s n="n"/></l><r>аав<s n="n"/></r></p></e><br />
<e><p><l>чуькле<s n="adj"/></l><r>бичкнv<s n="adj"/></r></p></e><br />
<e><p><l>мэхьтел<s n="adj"/></l><r>Удивленный<s n="adj"/></r></p></e><br />
<e><p><l>меслэхьэт<s n="v"/></l><r>зөвчлдх<s n="v"/></r></p></e><br />
<e><p><l>сийеь<s n="n"/></l><r>харллhн<s n="n"/></r></p></e><br />
<e><p><l>дервосе<s n="n"/></l><r>бәәшң<s n="n"/></r></p></e><br />
<e><p><l>мих<s n="n"/></l><r>хадасн<s n="n"/></r></p></e><br />
<e><p><l>ранжбе<s n="n"/></l><r>муҗг<s n="n"/></r></p></e><br />
<e><p><l>вежеьис<s n="v"/></l><r>биилх<s n="v"/></r></p></e><br />
</pre><br />
<br />
==xal==<br />
===morphology expansion===<br />
*verb: set form with ending -х as infinitive<br />
**<code>^нишх/ниш<v><inf>$^./.<sent>$</code><br />
*verb: added present tense suffix morphology -на/-нә, used for actions or conditions that are taking place or present at the time of speech<br />
**<code>^ицнә/иц<v><pres1>$^./.<sent>$</code><br />
*verb: added another form of present tense suffix morphology -чана/-чәнә, used for actions or conditions specifically in the dialogue<br />
**<code>^кевтчәнә/кевт<v><pres2>$^./.<sent>$</code><br />
*verb: added preterite suffix morphology -в, used for completed actions<br />
**<code>^гүүв/гүү<v><pret>$^./.<sent>$</code><br />
*added noun plurality suffix -нр, valid except after stems that end with р,г,с,к,н,ң <br />
**<code>^моhанр/моhа<n><pl><nom>$^./.<sent>$</code><br />
*added noun plurality suffix -муд/-мүд, valid except after stems that end with м , б , в , ң .<br />
**<code>^нумнмуд/нумн<n><pl><nom>$^./.<sent>$</code><br />
<br />
===twol rules===<br />
The rules in the twol file below the dotted line are rules added for this assignment.<br />
*rule: for the ablative case; realize archiphoneme {g} as г if stem ends with letter ң<br />
**<code>[PASS] Басң<n><sg><abl> => Басңгас</code><br />
*rule: for the accusative case; realize archiphoneme {I} as ы if stem ends in л, н, д, т, в.<br />
**<code>[PASS] мал<n><sg><acc> => малыг</code><br />
*rule: first part to prevent noun plural suffix нр from surfacing if stem ends in р, г, с, к, н, ң.<br />
*rule: second part to above rule<br />
**<code>така<n><pl><gen> => такамудин</code><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus: 33.60%<br />
*number of words in <code>large</code> corpus: 58060<br />
*number of stems in transducer: 304<br />
<br />
==jdt==<br />
===Morphology===<br />
I have added additional morphology to handle first, second, and third person copulas. <br />
<br />
===Disambiguation===<br />
I have added 4 new disambiguation rules. Ambiguity is now ~0.018<br />
<br />
* Remove adv if followed by a verb<br />
REMOVE (adv) IF (1 (v)) ;<br />
<br />
* Select verb tag if followed by a verb<br />
SELECT (v) IF (1 (v)) ;<br />
<br />
* Remove verb if not followed by a verb<br />
REMOVE (v) IF (NOT 1 (v)) ;<br />
<br />
example command:<br />
echo "темиз хуне" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" adv<br />
; "темиз" v REMOVE:17<br />
; "темиз" v sg p1 REMOVE:17<br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<.>"<br />
"." sent<br />
</pre><br />
example command:<br />
echo "темиз сох" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<темиз>"<br />
"темиз" v sg p1 SELECT:12<br />
"темиз" v SELECT:12<br />
; "темиз" adv REMOVE:11<br />
"<сох>"<br />
"сох" v<br />
"сох" v sg p1<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
* Remove conj if it is not the first word in the sentence<br />
REMOVE (conj) IF ((-1 (v)) OR (-1 (adj)) OR (-1 (n)) OR (-1 (adv)) OR (-1 (pr)) OR (-1 (det))) ;<br />
<br />
example command:<br />
echo "хуне егенер" | apertium -d . jdt-morph<br />
example output:<br />
<pre><br />
"<хуне>"<br />
"хун" n dat<br />
"хуне" n sg<br />
"<егенер>"<br />
"егенер" n sg<br />
; "егенер" conj REMOVE:19<br />
"<.>"<br />
"." sent<br />
</pre><br />
<br />
===Transducer Evaluation===<br />
*precision:<br />
*recall:<br />
*coverage over <code>large</code> corpus:<br />
*number of words in <code>large</code> corpus:<br />
*number of stems in transducer: 382<br />
<br />
==MT Evaluation==<br />
===jdt->xal===<br />
*WER over <code>longer</code> corpus: 100.00 %<br />
*PER over <code>longer</code> corpus: 100.00 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
===xal->jdt===<br />
*WER over <code>longer</code> corpus: 97.20 %<br />
*PER over <code>longer</code> corpus: 97.20 %<br />
*proportion of stems translated correctly in <code>longer</code> corpus:<br />
*coverage over <code>longer</code> corpus:<br />
*coverage over <code>large</code> corpus:<br />
*number of stems in <code>longer</code> corpus:<br />
*number of stems in <code>large</code> corpus:<br />
<br />
[[Category:Sp17 TranslationPairs]][[Category:Kalmyk]][[Category:Judeo-Tat]]</div>
Slim1