LING073 - User contributions [en]

User:Scasada1

2023-05-11T00:48:58Z

Scasada1: /* Overview */

== Sasha Casada ('24) ==

=== Overview ===

Hello. I am Sasha Casada ('24), a double major in Computer Science and Japanese. For LING073 (Spring 2023), I worked with [https://wikis.swarthmore.edu/ling073/User:Avykliu1 Alina Vykliuk '26] on Crimean Urum. You can find our work [https://wikis.swarthmore.edu/ling073/Urum here] as well as under the Urum language [https://wikis.swarthmore.edu/ling073/Category:Urum category] on the LING073 MediaWiki. You can also find my final project for the course [https://wikis.swarthmore.edu/ling073/User:Scasada1/Final_Project here]. Overall, I had a very nice time taking this class, definitely of my favorites for this year. Please feel free to use the information below to contact me with any questions!

=== Contact===
Please feel free to reach out to myself, or Alina, with any questions at: apertium (dot) urum (dot) project (at) gmail (dot) com, or at our Swarthmore email addresses. Thank you for reading.

[[Category: Urum]][[Category: sp23_students]]

User:Scasada1

2023-05-11T00:48:29Z

Scasada1: /* Overview */

=== Overview ===

Hello. I am Sasha Casada ('24), a double major in Computer Science and Japanese. For LING073 (Spring 2023), I worked with [https://wikis.swarthmore.edu/ling073/User:Avykliu1 Alina Vykliuk '26] on Crimean Urum. You can find our work [https://wikis.swarthmore.edu/ling073/Urum here] as well as under the Urum language [https://wikis.swarthmore.edu/ling073/Category:Urum category] on the LING073 MediaWiki. You can also find my final project for the course [https://wikis.swarthmore.edu/ling073/User:Scasada1/Final_Project here]. Overall, I had a very nice time taking this class, definitely of my favorites for this year. Please feel free to use the information below to contact me with any questions!

=== Contact===
Please feel free to reach out to myself, or Alina, with any questions at: apertium (dot) urum (dot) project (at) gmail (dot) com, or at our Swarthmore email addresses. Thank you for reading.

[[Category: Urum]][[Category: sp23_students]]

User:Scasada1

2023-05-11T00:46:49Z

Scasada1:

User:Scasada1

2023-05-11T00:46:20Z

Scasada1:

Hello. I am Sasha Casada ('24), a double major in Computer Science and Japanese. For LING073 (Spring 2023), I worked with [https://wikis.swarthmore.edu/ling073/User:Avykliu1 Alina Vykliuk '26] on Crimean Urum. You can find our work [https://wikis.swarthmore.edu/ling073/Urum here] as well as under the Urum language [https://wikis.swarthmore.edu/ling073/Category:Urum category] on this LING073 MediaWiki. Please feel free to reach out to myself, or Alina, with any questions at: apertium (dot) urum (dot) project (at) gmail (dot) com, or at our Swarthmore email addresses. Thank you for reading.

[[Category: Urum]][[Category: sp23_students]]

Urum and English/Contrastive Grammar

2023-05-11T00:34:45Z

Scasada1: /* (UUM ‣ ENG) Contrastive Grammar Evaluation */

== External resources ==

'''Transducer Links'''
* [https://github.com/apertium/apertium-eng English Transducer, Apertium] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum Source Code])
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum Transducer, LING073] ([https://github.com/scasada/apertium-eng.git Source Code])

'''Language Pair Repository'''
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum and English Translation Pairs] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-eng.git Source Code])

== (UUM ‣ ENG) Contrastive Grammar Evaluation ==

{{comment|Jonathan, I see your comments. Thank you for the feedback! I cannot work on addressing this right now (2 days left of finals), so perhaps I will revisit these corrections later during another academic year. Otherwise: hello to anyone viewing our wiki! -Scasada1}}

=== Example 1 (Verb Pronoun Marking) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a pronoun from the case of the verb.

* {{transferTest|uum|eng|алйым|I find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|sg}}|I{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алйых|we find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|pl}}|we{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алый|he finds}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|m}}{{tag|sg}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алыйлар|they find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|pl}}|they{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

=== Example 2 (Verb Past-Tense) ===

{{comment|There are spaces missing between the analyses. Also, I'm not sure this is a real grammar point, since the tag just gets copied? If it isn't, see if you can get one more grammar point. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the past-tense tags/markers.

* {{transferTest|uum|eng|булду|he found}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|алды|he stayed}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}stay{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|кӧрдӱ|he saw}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} see{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|хойду|he put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}put{{tag|vblex}}{{tag|past}}}}

=== Example 3 (Verb Negativity) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the negativity (in conjunction with past from the previous examples).

* {{transferTest|uum|eng|булмады|he did not find}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алмады|he did not stay}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} stay{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|кӧрмеди|he did not see}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} see{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|хоймады|he did not put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} put{{tag|vblex}}{{tag|pres}}}}

=== Example 4 (Ablative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|кӧктэн|from sky}}
*: {{transferMorphTest|uum|eng|кӧк{{tag|n}}{{tag|abl}}|from{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базардан|from market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|abl}}|from{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|завоттан|from factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|abl}}|from{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэн|from village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|abl}}|from{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

=== Example 5 (Locative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|бахчыда|in sky}}
*: {{transferMorphTest|uum|eng|бахчы{{tag|n}}{{tag|loc}}|in{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базарда|at market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|loc}}|at{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|бахчыда|at factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|loc}}|at{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэ|at village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|loc}}|at{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

[[Category:Sp23_ContrastiveGrammars]][[Category:Urum]][[Category:English]]

[[Category:Sp23_ContrastiveGrammars]] [[Category:Urum]] [[Category:English]]

Urum and English/Contrastive Grammar

2023-05-11T00:33:40Z

Scasada1: /* Example 5 (Locative Case) */

== External resources ==

'''Transducer Links'''
* [https://github.com/apertium/apertium-eng English Transducer, Apertium] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum Source Code])
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum Transducer, LING073] ([https://github.com/scasada/apertium-eng.git Source Code])

'''Language Pair Repository'''
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum and English Translation Pairs] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-eng.git Source Code])

== (UUM ‣ ENG) Contrastive Grammar Evaluation ==

=== Example 1 (Verb Pronoun Marking) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a pronoun from the case of the verb.

* {{transferTest|uum|eng|алйым|I find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|sg}}|I{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алйых|we find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|pl}}|we{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алый|he finds}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|m}}{{tag|sg}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алыйлар|they find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|pl}}|they{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

=== Example 2 (Verb Past-Tense) ===

{{comment|There are spaces missing between the analyses. Also, I'm not sure this is a real grammar point, since the tag just gets copied? If it isn't, see if you can get one more grammar point. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the past-tense tags/markers.

* {{transferTest|uum|eng|булду|he found}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|алды|he stayed}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}stay{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|кӧрдӱ|he saw}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} see{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|хойду|he put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}put{{tag|vblex}}{{tag|past}}}}

=== Example 3 (Verb Negativity) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the negativity (in conjunction with past from the previous examples).

* {{transferTest|uum|eng|булмады|he did not find}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алмады|he did not stay}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} stay{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|кӧрмеди|he did not see}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} see{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|хоймады|he did not put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} put{{tag|vblex}}{{tag|pres}}}}

=== Example 4 (Ablative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|кӧктэн|from sky}}
*: {{transferMorphTest|uum|eng|кӧк{{tag|n}}{{tag|abl}}|from{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базардан|from market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|abl}}|from{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|завоттан|from factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|abl}}|from{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэн|from village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|abl}}|from{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

=== Example 5 (Locative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|бахчыда|in sky}}
*: {{transferMorphTest|uum|eng|бахчы{{tag|n}}{{tag|loc}}|in{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базарда|at market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|loc}}|at{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|бахчыда|at factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|loc}}|at{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэ|at village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|loc}}|at{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

[[Category:Sp23_ContrastiveGrammars]][[Category:Urum]][[Category:English]]

[[Category:Sp23_ContrastiveGrammars]] [[Category:Urum]] [[Category:English]]

Urum and English/Contrastive Grammar

2023-05-11T00:33:33Z

Scasada1: /* Example 5 (Locative Case) */

== External resources ==

'''Transducer Links'''
* [https://github.com/apertium/apertium-eng English Transducer, Apertium] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum Source Code])
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum Transducer, LING073] ([https://github.com/scasada/apertium-eng.git Source Code])

'''Language Pair Repository'''
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum and English Translation Pairs] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-eng.git Source Code])

== (UUM ‣ ENG) Contrastive Grammar Evaluation ==

=== Example 1 (Verb Pronoun Marking) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a pronoun from the case of the verb.

* {{transferTest|uum|eng|алйым|I find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|sg}}|I{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алйых|we find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|pl}}|we{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алый|he finds}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|m}}{{tag|sg}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алыйлар|they find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|pl}}|they{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

=== Example 2 (Verb Past-Tense) ===

{{comment|There are spaces missing between the analyses. Also, I'm not sure this is a real grammar point, since the tag just gets copied? If it isn't, see if you can get one more grammar point. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the past-tense tags/markers.

* {{transferTest|uum|eng|булду|he found}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|алды|he stayed}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}stay{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|кӧрдӱ|he saw}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} see{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|хойду|he put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}put{{tag|vblex}}{{tag|past}}}}

=== Example 3 (Verb Negativity) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the negativity (in conjunction with past from the previous examples).

* {{transferTest|uum|eng|булмады|he did not find}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алмады|he did not stay}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} stay{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|кӧрмеди|he did not see}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} see{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|хоймады|he did not put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} put{{tag|vblex}}{{tag|pres}}}}

=== Example 4 (Ablative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|кӧктэн|from sky}}
*: {{transferMorphTest|uum|eng|кӧк{{tag|n}}{{tag|abl}}|from{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базардан|from market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|abl}}|from{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|завоттан|from factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|abl}}|from{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэн|from village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|abl}}|from{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

=== Example 5 (Locative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}
{{comment|Understood. I cannot work on this right now, perhaps I will revisit these corrections later. Hello to anyone viewing our wiki! -Scasada1}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|бахчыда|in sky}}
*: {{transferMorphTest|uum|eng|бахчы{{tag|n}}{{tag|loc}}|in{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базарда|at market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|loc}}|at{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|бахчыда|at factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|loc}}|at{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэ|at village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|loc}}|at{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

[[Category:Sp23_ContrastiveGrammars]][[Category:Urum]][[Category:English]]

[[Category:Sp23_ContrastiveGrammars]] [[Category:Urum]] [[Category:English]]

Urum and English/Contrastive Grammar

2023-05-11T00:32:10Z

Scasada1: /* Example 3 (Verb Negativity) */

== External resources ==

'''Transducer Links'''
* [https://github.com/apertium/apertium-eng English Transducer, Apertium] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum Source Code])
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum Transducer, LING073] ([https://github.com/scasada/apertium-eng.git Source Code])

'''Language Pair Repository'''
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum and English Translation Pairs] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-eng.git Source Code])

== (UUM ‣ ENG) Contrastive Grammar Evaluation ==

=== Example 1 (Verb Pronoun Marking) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a pronoun from the case of the verb.

* {{transferTest|uum|eng|алйым|I find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|sg}}|I{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алйых|we find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|pl}}|we{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алый|he finds}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|m}}{{tag|sg}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алыйлар|they find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|pl}}|they{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

=== Example 2 (Verb Past-Tense) ===

{{comment|There are spaces missing between the analyses. Also, I'm not sure this is a real grammar point, since the tag just gets copied? If it isn't, see if you can get one more grammar point. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the past-tense tags/markers.

* {{transferTest|uum|eng|булду|he found}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|алды|he stayed}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}stay{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|кӧрдӱ|he saw}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} see{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|хойду|he put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}put{{tag|vblex}}{{tag|past}}}}

=== Example 3 (Verb Negativity) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the negativity (in conjunction with past from the previous examples).

* {{transferTest|uum|eng|булмады|he did not find}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алмады|he did not stay}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} stay{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|кӧрмеди|he did not see}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} see{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|хоймады|he did not put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} put{{tag|vblex}}{{tag|pres}}}}

=== Example 4 (Ablative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|кӧктэн|from sky}}
*: {{transferMorphTest|uum|eng|кӧк{{tag|n}}{{tag|abl}}|from{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базардан|from market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|abl}}|from{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|завоттан|from factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|abl}}|from{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэн|from village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|abl}}|from{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

=== Example 5 (Locative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|бахчыда|in sky}}
*: {{transferMorphTest|uum|eng|бахчы{{tag|n}}{{tag|loc}}|in{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базарда|at market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|loc}}|at{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|бахчыда|at factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|loc}}|at{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэ|at village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|loc}}|at{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

[[Category:Sp23_ContrastiveGrammars]][[Category:Urum]][[Category:English]]

[[Category:Sp23_ContrastiveGrammars]] [[Category:Urum]] [[Category:English]]

Urum and English/Contrastive Grammar

2023-05-11T00:31:55Z

Scasada1: /* Example 3 (Verb Negativity) */

== External resources ==

'''Transducer Links'''
* [https://github.com/apertium/apertium-eng English Transducer, Apertium] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum Source Code])
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum Transducer, LING073] ([https://github.com/scasada/apertium-eng.git Source Code])

'''Language Pair Repository'''
* [https://wikis.swarthmore.edu/ling073/index.php?title=Urum_and_English Urum and English Translation Pairs] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-eng.git Source Code])

== (UUM ‣ ENG) Contrastive Grammar Evaluation ==

=== Example 1 (Verb Pronoun Marking) ===

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a pronoun from the case of the verb.

* {{transferTest|uum|eng|алйым|I find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|sg}}|I{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алйых|we find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p1}}{{tag|pl}}|we{{tag|prn}}{{tag|subj}}{{tag|p1}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алый|he finds}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|m}}{{tag|sg}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алыйлар|they find}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|pres}}{{tag|p3}}{{tag|pl}}|they{{tag|prn}}{{tag|subj}}{{tag|p3}}{{tag|mf}}{{tag|pl}}find{{tag|vblex}}{{tag|pres}}}}

=== Example 2 (Verb Past-Tense) ===

{{comment|There are spaces missing between the analyses. Also, I'm not sure this is a real grammar point, since the tag just gets copied? If it isn't, see if you can get one more grammar point. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the past-tense tags/markers.

* {{transferTest|uum|eng|булду|he found}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}find{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|алды|he stayed}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}stay{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|кӧрдӱ|he saw}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} see{{tag|vblex}}{{tag|past}}}}

* {{transferTest|uum|eng|хойду|he put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}}put{{tag|vblex}}{{tag|past}}}}

=== Example 3 (Verb Negativity) ===

{{comment|Missing spaces as above. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to extract the negativity (in conjunction with past from the previous examples).

* {{transferTest|uum|eng|булмады|he did not find}}
*: {{transferMorphTest|uum|eng|бул{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} find{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|алмады|he did not stay}}
*: {{transferMorphTest|uum|eng|ал{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} stay{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|кӧрмеди|he did not see}}
*: {{transferMorphTest|uum|eng|кӧр{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} see{{tag|vblex}}{{tag|pres}}}}

* {{transferTest|uum|eng|хоймады|he did not put}}
*: {{transferMorphTest|uum|eng|хой{{tag|v}}{{tag|tv}}{{tag|neg}}{{tag|past}}{{tag|p3}}{{tag|sg}}|he{{tag|prn}}{{tag|p3}}{{tag|m}}{{tag|sg}} did{{tag|vblex}}{{tag|past}} not{{tag|adv}} put{{tag|vblex}}{{tag|pres}}}}

=== Example 4 (Ablative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|кӧктэн|from sky}}
*: {{transferMorphTest|uum|eng|кӧк{{tag|n}}{{tag|abl}}|from{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базардан|from market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|abl}}|from{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|завоттан|from factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|abl}}|from{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэн|from village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|abl}}|from{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

=== Example 5 (Locative Case) ===

{{comment|Missing spaces as above. The description doesn't match the phenomenon. -JNW}}

In Urum, verbs are agglutinative. In order to translate correctly into English, we need to create a preposition from the case of the verb.

* {{transferTest|uum|eng|бахчыда|in sky}}
*: {{transferMorphTest|uum|eng|бахчы{{tag|n}}{{tag|loc}}|in{{tag|pr}}sky{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|базарда|at market}}
*: {{transferMorphTest|uum|eng|базар{{tag|n}}{{tag|loc}}|at{{tag|pr}}market{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|бахчыда|at factory}}
*: {{transferMorphTest|uum|eng|завот{{tag|n}}{{tag|loc}}|at{{tag|pr}}factory{{tag|n}}{{tag|sg}}}}

* {{transferTest|uum|eng|кӧйдэ|at village}}
*: {{transferMorphTest|uum|eng|кӧй{{tag|n}}{{tag|loc}}|at{{tag|pr}}village{{tag|n}}{{tag|sg}}}}

[[Category:Sp23_ContrastiveGrammars]][[Category:Urum]][[Category:English]]

[[Category:Sp23_ContrastiveGrammars]] [[Category:Urum]] [[Category:English]]

Urum

2023-05-11T00:30:06Z

Scasada1: /* Developed Resources */

Resource documentation for Urum [uˈrum].

Urum is a language spoken in both Ukraine, and Georgia by an [https://www.ethnologue.com/18/language/uum/ estimated 190,000 (2000) speakers]. It is a Turkic language with the specification of the language family being as follows: [Turkic → Common Turkic → Kipchak → West Kypchak]. Urum has two distinct dialects: [http://projects.turkmas.uoa.gr/urum/gcaucasianurum.html Caucasian Urum] (local to Greece and the surrounding country, utilizing the Greek alphabet; it is also important to note that the variety of Urum called: Turkish Urum, is considered to be a variety of this), and [https://web.archive.org/web/20080127062418/http://www.publish.diaspora.ru/magazin/articles/russia026_1.shtml Ukranian Urum] (also known as Greek-Tatar; utilizing the Cyrillic alphabet).

== Developed Resources ==

* [https://wikis.swarthmore.edu/ling073/Urum/Keyboard Urum Keyboard Layout] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-keyboard Source Code])
* [https://github.swarthmore.edu/Ling073-sp23/ling073-uum-tools Urum Language Tools Respository]
* [https://wikis.swarthmore.edu/ling073/Urum Urum Apertium Transducer] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum.git Source Code])
* [https://wikis.swarthmore.edu/ling073/Urum_and_English Urum Apertium RBMT System] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-eng.git Source Code])

== Developed Documentation ==

* [https://wikis.swarthmore.edu/ling073/Urum/Disambiguation Urum Apertium Pipeline Disambiguation]
* [https://wikis.swarthmore.edu/ling073/Urum/Grammar Urum Language Grammar Overview]

== External Resources ==

=== Websites ===

* [https://en.wikipedia.org/wiki/Urum_language Wikipedia: Urum Language]

* [https://www.endangeredlanguages.com/lang/3004/samples/10744 Endangered Languages: Urum] (this article is a stub, perhaps we can improve it!)

* [https://en.wiktionary.org/wiki/Urum Wiktionary: Urum Article] (this article is a stub, perhaps we can improve it!)

* [https://vlo.clarin.eu/record/https_58__47__47_hdl.handle.net_47_1839_47_00-0000-0000-0021-4DA5-8;jsessionid=2F3E6F152EBD6D683AF31EDF4F3C54CE?0 Virtual Language Observatory: Urum] (this article is a stub, perhaps we can improve it!)

=== Computational Resources ===

* [https://glosbe.com/en/uum Globse: Online Urum Translator] (perhaps poor in quality, uses algorithmic analysis)

=== Dictionaries ===

* [http://old.unesco.kz/qypchaq/Urum_Dictionary.htm Urum Dictionary Online] Олександр Гаркавець. Урумський словник – Алмати: Баур, 2000.– 632 с. (Oleksandr Harkavets. Urum Dictionary. Almaty: Baur, 2000. – 632 p.) (Rights Reserved)

=== Audio/Video ===

* [https://www.youtube.com/watch?v=qHdrkdz27og&t=1s Urum Alphabet in Verses] (Publicly Available on YouTube)

* [https://www.youtube.com/@urum6724 Урум дыли: Урумська: Urum] (YouTube content, more available at the link)

* [https://obuchalka.org/20200630122200/urumskii-yazik-urum-dili-priazovskii-variant-uchebnoe-posobie-dlya-nachinauschih-s-audioprilojeniem-smolina-m-2008.html Урумский язык, Урум дили (приазовский вариант), Учебное пособие для начинающих с аудиоприложением, Смолина М., 2008.] (The Urum language (Azov region version), Tutorial for beginners with audio tapes, Smolina M., 2008)

=== Grammatical Descriptions ===

* [https://core.ac.uk/download/pdf/211853704.pdf Caucasian Urums and Urum language (Kafkasya Urumları ve Urum Dili)] by Stavros Skopeteas, Bielefeld University, submitted to the ''Handbook of Endangered Turkic Languages'', ed. by Süer Eker and Ülkü Çelik

* [http://projects.turkmas.uoa.gr/urum/ Urum Documentation Project] - A cooperative research project at the Universities of Athens, Bielefeld, Bremen, and Potsdam (funded by the Latsis Foundation). This is a multipart work, which contains multiple sections on various aspects of the Urum language, which was all compiled together by the same group of researchers working in conjunction to create a comprehensive academic, and mostly cursory, description of the Caucasian variety of the language. (Rights Reserved; Can Gain Permission of the Researching Body)
** [http://projects.turkmas.uoa.gr/urum/w_index.html Description of Basic Lexicon]
** [http://projects.turkmas.uoa.gr/urum/t_index.html Narrative Collection]
** [http://projects.turkmas.uoa.gr/urum/bharmony.html Vowel Harmony and Noun Inflections]
** [http://projects.turkmas.uoa.gr/urum/s_index.html Basic Grammatical Structures]

=== Scientific Works ===

* [https://core.ac.uk/download/pdf/211853704.pdf Caucasian Urums and Urum language (Kafkasya Urumları ve Urum Dili)] Stavros Skopeteas, Bielefeld University; from ''Handbook of Endangered Turkic Languages'' ed. by Süer Eker and Ülkü Çelik (Rights Reserved)
** '''Covers Topics/Provides Resources On''': Orthography, Phonology, Word Classes and Morphological Categories, Lexicon; Sample Texts

* 'The Syntax of Focus in Caucasian Urum' by Stefanie Schröter (Rights Reserved)
* 'From Double to Dependent Marking? An Investigation of Possessive Constructions in Caucasian Urum' by Johanna Neugebauer in ''DE GRUYTER MOUTON'' (STUF 2016; 69(2): 277–307) (Rights Reserved)
** Covers the results of study on possessive constructions in the Turkish variation of Caucasian Urum
* 'Discovering the Prehistory of Multilingual Situations in the Lexicon, an Empirical Study on the Caucasian Urum Vocabulary' - Veronika Ries, Stavros Skopeteas, Emrah Turan, Kirstin Nahrmann (Bielefeld) (Rights Reserved)
* "Языки мира. Тюркские языки." Российская академия наук, институт языкознания. Издательский дом, Кыргызстан, Бишкек. ("Languages of the world. Turkic languages." Russian Academy of Sciences. Kyrgyzstan, Bishkek)

=== Text Resources/Corpora ===

* UKRAYNA'DAK‹ URUM TÜRKLER‹ VE FOLKLORU (Urum Turks and Urum's folklore in Ukraine) - Doç. Dr. Hülya KASAPO⁄LU ÇENGEL* (Rights Reserved)
* Хара Гавриил Спиридонович - Урумский сборник из Гранитного (Карани) Тельмановского района Донецкой области (Hara Gavriil Spiridonovich - Urum Collection from Granite (Karani) Telmanovsky District of Donetsk Region) - (ANCIENT TEXT, BARELY CAN READ! PERHAPS WE COULD USE SOME KIND OF IMAGE/PDF OCR PROGRAM!)
* советская археология 1973 (Soviet archeology 1973) (NOT SURE HOW TO READ THIS...)
* Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я (Oleksandr HARKAVETS THE URUM PEOPLE OF NORTH AZOV)

== Credits ==
This wiki was created by Sasha Casada ('24) [[User:Scasada1]] and Alina Vykliuk ('26) [[User:avykliu1]]

[[Category:Urum]]
[[Category:sp23_ResourceDocumentation]]

Urum

2023-05-11T00:29:49Z

Scasada1: /* Developed Resources */

Resource documentation for Urum [uˈrum].

Urum is a language spoken in both Ukraine, and Georgia by an [https://www.ethnologue.com/18/language/uum/ estimated 190,000 (2000) speakers]. It is a Turkic language with the specification of the language family being as follows: [Turkic → Common Turkic → Kipchak → West Kypchak]. Urum has two distinct dialects: [http://projects.turkmas.uoa.gr/urum/gcaucasianurum.html Caucasian Urum] (local to Greece and the surrounding country, utilizing the Greek alphabet; it is also important to note that the variety of Urum called: Turkish Urum, is considered to be a variety of this), and [https://web.archive.org/web/20080127062418/http://www.publish.diaspora.ru/magazin/articles/russia026_1.shtml Ukranian Urum] (also known as Greek-Tatar; utilizing the Cyrillic alphabet).

== Developed Resources ==

* [https://wikis.swarthmore.edu/ling073/Urum/Keyboard Urum Keyboard Layout] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-keyboard Source Code])
* [https://wikis.swarthmore.edu/ling073/Urum Urum Apertium Transducer] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum.git Source Code])
* [https://wikis.swarthmore.edu/ling073/Urum_and_English Urum Apertium RBMT System] ([https://github.swarthmore.edu/Ling073-sp23/ling073-uum-eng.git Source Code])
* [https://github.swarthmore.edu/Ling073-sp23/ling073-uum-tools Urum Language Tools Respository]

== Developed Documentation ==

* [https://wikis.swarthmore.edu/ling073/Urum/Disambiguation Urum Apertium Pipeline Disambiguation]
* [https://wikis.swarthmore.edu/ling073/Urum/Grammar Urum Language Grammar Overview]

== External Resources ==

=== Websites ===

* [https://en.wikipedia.org/wiki/Urum_language Wikipedia: Urum Language]

* [https://www.endangeredlanguages.com/lang/3004/samples/10744 Endangered Languages: Urum] (this article is a stub, perhaps we can improve it!)

* [https://en.wiktionary.org/wiki/Urum Wiktionary: Urum Article] (this article is a stub, perhaps we can improve it!)

* [https://vlo.clarin.eu/record/https_58__47__47_hdl.handle.net_47_1839_47_00-0000-0000-0021-4DA5-8;jsessionid=2F3E6F152EBD6D683AF31EDF4F3C54CE?0 Virtual Language Observatory: Urum] (this article is a stub, perhaps we can improve it!)

=== Computational Resources ===

* [https://glosbe.com/en/uum Globse: Online Urum Translator] (perhaps poor in quality, uses algorithmic analysis)

=== Dictionaries ===

* [http://old.unesco.kz/qypchaq/Urum_Dictionary.htm Urum Dictionary Online] Олександр Гаркавець. Урумський словник – Алмати: Баур, 2000.– 632 с. (Oleksandr Harkavets. Urum Dictionary. Almaty: Baur, 2000. – 632 p.) (Rights Reserved)

=== Audio/Video ===

* [https://www.youtube.com/watch?v=qHdrkdz27og&t=1s Urum Alphabet in Verses] (Publicly Available on YouTube)

* [https://www.youtube.com/@urum6724 Урум дыли: Урумська: Urum] (YouTube content, more available at the link)

* [https://obuchalka.org/20200630122200/urumskii-yazik-urum-dili-priazovskii-variant-uchebnoe-posobie-dlya-nachinauschih-s-audioprilojeniem-smolina-m-2008.html Урумский язык, Урум дили (приазовский вариант), Учебное пособие для начинающих с аудиоприложением, Смолина М., 2008.] (The Urum language (Azov region version), Tutorial for beginners with audio tapes, Smolina M., 2008)

=== Grammatical Descriptions ===

* [https://core.ac.uk/download/pdf/211853704.pdf Caucasian Urums and Urum language (Kafkasya Urumları ve Urum Dili)] by Stavros Skopeteas, Bielefeld University, submitted to the ''Handbook of Endangered Turkic Languages'', ed. by Süer Eker and Ülkü Çelik

* [http://projects.turkmas.uoa.gr/urum/ Urum Documentation Project] - A cooperative research project at the Universities of Athens, Bielefeld, Bremen, and Potsdam (funded by the Latsis Foundation). This is a multipart work, which contains multiple sections on various aspects of the Urum language, which was all compiled together by the same group of researchers working in conjunction to create a comprehensive academic, and mostly cursory, description of the Caucasian variety of the language. (Rights Reserved; Can Gain Permission of the Researching Body)
** [http://projects.turkmas.uoa.gr/urum/w_index.html Description of Basic Lexicon]
** [http://projects.turkmas.uoa.gr/urum/t_index.html Narrative Collection]
** [http://projects.turkmas.uoa.gr/urum/bharmony.html Vowel Harmony and Noun Inflections]
** [http://projects.turkmas.uoa.gr/urum/s_index.html Basic Grammatical Structures]

=== Scientific Works ===

* [https://core.ac.uk/download/pdf/211853704.pdf Caucasian Urums and Urum language (Kafkasya Urumları ve Urum Dili)] Stavros Skopeteas, Bielefeld University; from ''Handbook of Endangered Turkic Languages'' ed. by Süer Eker and Ülkü Çelik (Rights Reserved)
** '''Covers Topics/Provides Resources On''': Orthography, Phonology, Word Classes and Morphological Categories, Lexicon; Sample Texts

* 'The Syntax of Focus in Caucasian Urum' by Stefanie Schröter (Rights Reserved)
* 'From Double to Dependent Marking? An Investigation of Possessive Constructions in Caucasian Urum' by Johanna Neugebauer in ''DE GRUYTER MOUTON'' (STUF 2016; 69(2): 277–307) (Rights Reserved)
** Covers the results of study on possessive constructions in the Turkish variation of Caucasian Urum
* 'Discovering the Prehistory of Multilingual Situations in the Lexicon, an Empirical Study on the Caucasian Urum Vocabulary' - Veronika Ries, Stavros Skopeteas, Emrah Turan, Kirstin Nahrmann (Bielefeld) (Rights Reserved)
* "Языки мира. Тюркские языки." Российская академия наук, институт языкознания. Издательский дом, Кыргызстан, Бишкек. ("Languages of the world. Turkic languages." Russian Academy of Sciences. Kyrgyzstan, Bishkek)

=== Text Resources/Corpora ===

* UKRAYNA'DAK‹ URUM TÜRKLER‹ VE FOLKLORU (Urum Turks and Urum's folklore in Ukraine) - Doç. Dr. Hülya KASAPO⁄LU ÇENGEL* (Rights Reserved)
* Хара Гавриил Спиридонович - Урумский сборник из Гранитного (Карани) Тельмановского района Донецкой области (Hara Gavriil Spiridonovich - Urum Collection from Granite (Karani) Telmanovsky District of Donetsk Region) - (ANCIENT TEXT, BARELY CAN READ! PERHAPS WE COULD USE SOME KIND OF IMAGE/PDF OCR PROGRAM!)
* советская археология 1973 (Soviet archeology 1973) (NOT SURE HOW TO READ THIS...)
* Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я (Oleksandr HARKAVETS THE URUM PEOPLE OF NORTH AZOV)

== Credits ==
This wiki was created by Sasha Casada ('24) [[User:Scasada1]] and Alina Vykliuk ('26) [[User:avykliu1]]

[[Category:Urum]]
[[Category:sp23_ResourceDocumentation]]

User:Avykliu1/Final Project

2023-05-11T00:27:57Z

Scasada1:

Please see [https://wikis.swarthmore.edu/ling073/User:Scasada1/Final_Project here] for our (avykliu1 & scasada1) final project.

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Avykliu1/Final Project

2023-05-11T00:27:45Z

Scasada1: Created page with "Please see: [https://wikis.swarthmore.edu/ling073/User:Scasada1/Final_Project this page] for our (avykliu1 & scasada1) final project. Category:Urum Category:Sp23_FinalP..."

Please see: [https://wikis.swarthmore.edu/ling073/User:Scasada1/Final_Project this page] for our (avykliu1 & scasada1) final project.

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:25:48Z

Scasada1: /* Selected Example Sentences from the Large Corpus */

User:Scasada1/Final Project

2023-05-11T00:25:02Z

Scasada1: /* Previous Work */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Urum Urum Transducer]

[https://wikis.swarthmore.edu/ling073/Urum_and_English Urum-English Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (uum ‣ eng) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

=== RBMT Evalutation ===

{| class="wikitable"
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

== Future Work ==
* We currently have a very rich analysis system, but we need to expand our current dictionary significantly (~1000+ words, ideally)
** In order to solve this, we have a publicly available spreadsheet dictionary for Urum, based off a Ukranian-Urum dictionary translated to English ([https://docs.google.com/spreadsheets/d/1AowA3a4iWiNgHLxARw0QwWXgNq1_aegIrjV7eHxO9Xc/edit?usp=sharing which is available on Google Sheets]).
* Refine our Apertium (uum ‣ eng) transfer pipeline; i.e, robust expansion of lexical selection and lexical transfer as issues begin to arise.

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:24:00Z

Scasada1: /* Links to Public Git Repo and Poster */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (uum ‣ eng) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

=== RBMT Evalutation ===

{| class="wikitable"
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

== Future Work ==
* We currently have a very rich analysis system, but we need to expand our current dictionary significantly (~1000+ words, ideally)
** In order to solve this, we have a publicly available spreadsheet dictionary for Urum, based off a Ukranian-Urum dictionary translated to English ([https://docs.google.com/spreadsheets/d/1AowA3a4iWiNgHLxARw0QwWXgNq1_aegIrjV7eHxO9Xc/edit?usp=sharing which is available on Google Sheets]).
* Refine our Apertium (uum ‣ eng) transfer pipeline; i.e, robust expansion of lexical selection and lexical transfer as issues begin to arise.

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:23:04Z

Scasada1: /* Future Work */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (uum ‣ eng) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

=== RBMT Evalutation ===

{| class="wikitable"
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

== Future Work ==
* We currently have a very rich analysis system, but we need to expand our current dictionary significantly (~1000+ words, ideally)
** In order to solve this, we have a publicly available spreadsheet dictionary for Urum, based off a Ukranian-Urum dictionary translated to English ([https://docs.google.com/spreadsheets/d/1AowA3a4iWiNgHLxARw0QwWXgNq1_aegIrjV7eHxO9Xc/edit?usp=sharing which is available on Google Sheets]).
* Refine our Apertium (uum ‣ eng) transfer pipeline; i.e, robust expansion of lexical selection and lexical transfer as issues begin to arise.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:21:36Z

Scasada1: /* Future Work */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (uum ‣ eng) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

=== RBMT Evalutation ===

{| class="wikitable"
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

== Future Work ==
* We currently have a very rich analysis system, but we need to expand our current dictionary significantly (~1000+ words, ideally)
** In order to solve this, we have a publicly available spreadsheet dictionary for Urum, based off a Ukranian-Urum dictionary translated to English (which is available on GitHub).
* Refine our Apertium (uum ‣ eng) transfer pipeline; i.e, robust expansion of lexical selection and lexical transfer as issues begin to arise.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:20:54Z

Scasada1: /* Future Work */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (uum ‣ eng) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

=== RBMT Evalutation ===

{| class="wikitable"
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

== Future Work ==
* We currently have a very rich analysis system, but we need to expand our current dictionary significantly (~1000+ words, ideally)
** In order to solve this, we have a publicly available spreadsheet dictionary for Urum, based off a Ukranian-Urum dictionary translated to English (which is available on GitHub).
* Refine our Apertium

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:18:50Z

Scasada1: /* Final Evaluation */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (uum ‣ eng) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

=== RBMT Evalutation ===

{| class="wikitable"
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:18:26Z

Scasada1: /* Final Evaluation */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (uum ‣ eng) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:18:10Z

Scasada1: /* Transfer Overview (UUM ‣ ENG) */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (uum ‣ eng) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:17:59Z

Scasada1: /* Transducer Overview */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer Overview (uum) ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (UUM ‣ ENG) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:17:50Z

Scasada1: /* Transfer Overview */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer Overview ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview (UUM ‣ ENG) ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:17:21Z

Scasada1: /* Transfer Overview */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer Overview ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:17:08Z

Scasada1: /* Transfer Overview */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer Overview ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%
* Bilingual Dictionary Entries: 260
* Total LRX Rules: 8

(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:16:09Z

Scasada1: /* Machine Translation */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer Overview ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Transfer Overview ===
*Urum Bible: "Nativity":
** WER: 81.29%
** PER: 57.19%

(The amount of error present here is likely due to the type of language used in the King James translation of the bible, as we used this as gold for the English translation. We believe that the word-error rate, and more specifically PER, would be far lower with a version of the bible in English which is more accurate to the original text).

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:12:43Z

Scasada1: /* Transducer Overview */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer Overview ===
*Coverage over the "Urums of North Azov: Collection of Narratives": 60.071%
*Precision: 55.678%
*Recall: 73.077%
*Total Stems: 471
**N-Roots: 121
**V-Roots: 61
**Adv/Adj-Roots: 22

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-11T00:11:25Z

Scasada1: /* Transducer */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer Overview ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-10T21:33:09Z

Scasada1: /* Final Evaluation */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% (163/170)
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% (53K/186K)
|N/A
|}

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-10T21:32:47Z

Scasada1: /* Final Evaluation */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

{| class="wikitable"
!colspan="6"|RBMT Evaluation
|-
!'''Corpus'''
!'''Tokens'''
!'''Coverage'''
!'''Accuracy'''
|-
|Urum Bible: "Nativity"
|170
|95.882% ''(163/170)''
|91.765%
|-
|"Urums of North Azov: Collection of Narratives"
|18K
|28.487% ''(53K/186K)''
|N/A
|}

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-10T21:23:01Z

Scasada1: /* Nuosu-Mandarin Machine Translation */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Urum-English Machine Translation ===
* Added one structural transfer rule to with the formation of NPs in the case of proper nouns (prior not realizing correctly, omitting 'the' and instead having no output, ex: "king #Herod"
* Added 100+ new lexicon items to our bilingual dictionary, making the total ~250.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-10T21:21:08Z

Scasada1: /* Urum Transducer */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)
* We added a large number of additional .lexd PATTERNS including:
**''VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, & CaseTerm''

=== Nuosu-Mandarin Machine Translation ===
* Added one structural transfer rule to deal with sentence level particles in Nuosu that are omitted in Mandarin.
* Added one structural transfer rule to insert Mandarin particle 'de', which can function as predicate marker, possesive marker, adjectivizer, etc.
* Added ~80 new lexicon to our bilingual dictionary, making the total ~200.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-10T21:18:25Z

Scasada1: /* Additions */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Urum Transducer ===
* We added 100+ stems to the .lexd file, covering common nouns, verbs, adjectives, proper nouns, etc. as they occurred in terms of frequency in appearance within (Олександр ГАРКАВЕЦЬ УРУМИ НАДАЗОВ’Я)
* We added a large number of additional .lexd PATTERNS including: VFutureDefinite, VFutureIndefinite, VTransgressive, VPastParticiple, NumOrdinal, NumDiscrete, CaseIns, CaseAbe, and CaseTerm.
* Expanded our number of rules in .twol to cover the rest of Urum's vowel and consonant harmony: (Total Rules: 16 → 24)

=== Nuosu-Mandarin Machine Translation ===
* Added one structural transfer rule to deal with sentence level particles in Nuosu that are omitted in Mandarin.
* Added one structural transfer rule to insert Mandarin particle 'de', which can function as predicate marker, possesive marker, adjectivizer, etc.
* Added ~80 new lexicon to our bilingual dictionary, making the total ~200.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-10T21:09:44Z

Scasada1: /* Final Project Description */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|800px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Nuosu Transducer ===
* We added 100+ stems to the lexd. file, covering common nouns, verbs, and adjectives mentioned in A Grammar of Nuosu (Gerner 2013).
* We attempted to implement the negation suffix 'ap' using twol rule; currently the negation infix is only working for monosyllabic verbs and adjectives.

=== Nuosu-Mandarin Machine Translation ===
* Added one structural transfer rule to deal with sentence level particles in Nuosu that are omitted in Mandarin.
* Added one structural transfer rule to insert Mandarin particle 'de', which can function as predicate marker, possesive marker, adjectivizer, etc.
* Added ~80 new lexicon to our bilingual dictionary, making the total ~200.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-10T21:09:15Z

Scasada1: /* Final Project Description */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|1000px|thumb|right|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Nuosu Transducer ===
* We added 100+ stems to the lexd. file, covering common nouns, verbs, and adjectives mentioned in A Grammar of Nuosu (Gerner 2013).
* We attempted to implement the negation suffix 'ap' using twol rule; currently the negation infix is only working for monosyllabic verbs and adjectives.

=== Nuosu-Mandarin Machine Translation ===
* Added one structural transfer rule to deal with sentence level particles in Nuosu that are omitted in Mandarin.
* Added one structural transfer rule to insert Mandarin particle 'de', which can function as predicate marker, possesive marker, adjectivizer, etc.
* Added ~80 new lexicon to our bilingual dictionary, making the total ~200.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-10T21:08:53Z

Scasada1: /* Final Project Description */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work will consist of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, MT, text processing, etc.

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)] and [https://github.com/apertium-uum/sp2023-uum-poster/tree/main here] you may separately view a digital copy of our final presentation poster as it exists on GitHub.
[[File:Output-onlinepngtools (1).png|1000px|thumb|left|SP2023 Urum "Polished RBMT System for Crimean Urum" Poster]]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Nuosu Transducer ===
* We added 100+ stems to the lexd. file, covering common nouns, verbs, and adjectives mentioned in A Grammar of Nuosu (Gerner 2013).
* We attempted to implement the negation suffix 'ap' using twol rule; currently the negation infix is only working for monosyllabic verbs and adjectives.

=== Nuosu-Mandarin Machine Translation ===
* Added one structural transfer rule to deal with sentence level particles in Nuosu that are omitted in Mandarin.
* Added one structural transfer rule to insert Mandarin particle 'de', which can function as predicate marker, possesive marker, adjectivizer, etc.
* Added ~80 new lexicon to our bilingual dictionary, making the total ~200.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

File:Output-onlinepngtools (1).png

2023-05-10T21:06:42Z

Scasada1: Urum presentation poster for SP2023

Urum presentation poster for SP2023

File:Urum Pres.png

2023-05-10T21:05:33Z

Scasada1: Scasada1 uploaded a new version of File:Urum Pres.png

Urum presentation poster for SP2023

File:Urum Pres.png

2023-05-10T21:01:28Z

Scasada1: Urum presentation poster for SP2023

Urum presentation poster for SP2023

User:Scasada1/Final Project

2023-05-09T15:56:31Z

Scasada1: /* Final Project Description */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work consists of many parts, but we will outline our base goals here:
* Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
* *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, etc. (ADD INFORMATION FROM THE POSTER HERE!) :)

This will be the information that is provided on our poster

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Nuosu Transducer ===
* We added 100+ stems to the lexd. file, covering common nouns, verbs, and adjectives mentioned in A Grammar of Nuosu (Gerner 2013).
* We attempted to implement the negation suffix 'ap' using twol rule; currently the negation infix is only working for monosyllabic verbs and adjectives.

=== Nuosu-Mandarin Machine Translation ===
* Added one structural transfer rule to deal with sentence level particles in Nuosu that are omitted in Mandarin.
* Added one structural transfer rule to insert Mandarin particle 'de', which can function as predicate marker, possesive marker, adjectivizer, etc.
* Added ~80 new lexicon to our bilingual dictionary, making the total ~200.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-09T15:56:16Z

Scasada1: /* Final Project Description */

== Final Project Description ==
For our final project, we decided to work on a two-part expansion of our RBMT-system for our language (Crimean Urum). This work consists of many parts, but we will outline our base goals here:
- Improve our morphological transducer for Crimean Urum, aiming for a corpus coverage above 80%.
- *Improve our Crimean Urum machine translation accuracy.
We believe that our morphological transducer and machine translation system will be potentially helpful for living Urum language communities with tasks like generating novel sentences, and/or helping with an auto-correction system, etc. (ADD INFORMATION FROM THE POSTER HERE!) :)

This will be the information that is provided on our poster

We have released information for our project publicly on [https://github.com/apertium-uum GitHub (apertium-uum)]

== Previous Work ==

[https://wikis.swarthmore.edu/ling073/Nuosu/Transducer Nuosu Transducer]

[https://wikis.swarthmore.edu/ling073/Nuosu_and_Mandarin/Structural_transfer Nuosu-Mandarin Machine Translation]

== Additions ==

=== Nuosu Transducer ===
* We added 100+ stems to the lexd. file, covering common nouns, verbs, and adjectives mentioned in A Grammar of Nuosu (Gerner 2013).
* We attempted to implement the negation suffix 'ap' using twol rule; currently the negation infix is only working for monosyllabic verbs and adjectives.

=== Nuosu-Mandarin Machine Translation ===
* Added one structural transfer rule to deal with sentence level particles in Nuosu that are omitted in Mandarin.
* Added one structural transfer rule to insert Mandarin particle 'de', which can function as predicate marker, possesive marker, adjectivizer, etc.
* Added ~80 new lexicon to our bilingual dictionary, making the total ~200.

=== Corpus Construction ===
* Constructed small corpus with 30 sentences; the lexicon in this corpus are mostly added to our transducer and the bilingual dictionary, which is useful for evaluateing how the effectiveness of the structural transfer rules when the lexicon are mostly known.

* Constructed large corpus with 100 sentences; the sentences in this corpus are randomly selected from the Bible and the Grammar book, and evaluating over this corpus shows the general accuracy of our machine translation system.

==== Selected Example Sentences from the Large Corpus ====
* {{transferTest|iii|zho|ꋌꊋꋠꃅꀥ|他用力地跑}}
*: {{transferMorphTest|iii|zho|ꋌ{{tag|prn}} ꊋꋠ{{tag|adj}} ꃅ{{tag|advl}} ꀥ{{tag|vblex}}| 他{{tag|prn}} 用力地{{tag|adv}} 跑{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉪꊇꀃꑍꏦꎫꁧ|我们今天去街上}}
*: {{transferMorphTest|iii|zho|ꉪꊇ{{tag|prn}} ꀃꑍ{{tag|n}} ꏦꎫ{{tag|n}} ꁧ{{tag|vblex}}| 我们{{tag|prn}} 今天{{tag|n}} 去{{tag|vblex}} 街上{{tag|n}}}}

* {{transferTest|iii|zho|ꉢꎧꀋꅝꀐ|我没有喝红酒}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꎧ{{tag|n}} ꀋ{{tag|neg}} ꅝ{{tag|vblex}} ꀐ{{tag|part}}| 我{{tag|prn}} 没有{{tag|neg}} 喝{{tag|vblex}} 红酒{{tag|n}}}}

* {{transferTest|iii|zho|ꊾꆹꉬꇮꐯꀋꇊꌡ|所有人差不多地像}}
*: {{transferMorphTest|iii|zho|ꊾ{{tag|n}} ꆹ{{tag|part}} ꉬꇮ{{tag|adj}} ꐯ{{tag|RECL}} ꀋꇊ{{tag|adv}} ꌡ{{tag|vblex}}| 所有{{tag|adj}} 人{{tag|n}} 差不多地{{tag|adv}} 像{{tag|vblex}}}}

* {{transferTest|iii|zho|ꉢꃀꌠꀐ|我是老人了}}
*: {{transferMorphTest|iii|zho|ꉢ{{tag|prn}} ꃀꌠ{{tag|n}} ꀐ{{tag|part}} | 我{{tag|prn}} 现在{{tag|adv}} 是{{tag|pr}} 老人{{tag|n}}}}

== Final Evaluation ==

=== Transducer ===
*Coverage over the Bible: 78.5%
*Precision: 97.60%
*Recall: 86.46%

=== Machine Translation ===
*Small Corpora:
** WER: 37.9%
** PER: 17.8%
*Large Corpora:
** WER: 107.5%
** PER: 80.9%
(We think ths difference in the WERs and PERs is due to the large number of unknown nouns and pronouns contained in the bible that are not added to our bilingual dictionary; also the reference text are polished translations selected from the bible but the test text are direct translations using our machine translation system that maps between lexicons)

== Future Work ==
* Add more structural transfer rules to deal with compounding in Mandarin.
* Reduce twol rule conflicts in the transducer; improve on the implementation of the negation suffix 'ap'.
* Enhance translation accuracy in collaboration with native speakers.

== Links to Public Git Repo and Poster ==

*Morphological Analyser, generator, and disambiguator for Nuosu
**[https://github.com/AgnesWang000/Nuosu-Transducer Morphological Transducer]

*Machine Translation between Nuosu and Mandarin
**[https://github.com/AgnesWang000/Nuosu-Mandarin-RBMT Machine Translation System]

*Improved Mandarin Transducer
**[https://github.com/AgnesWang000/Apertium-zho Apertium-zho]

*Summary Poster of the Project
**[https://docs.google.com/presentation/d/1qZwlHxp997AG8Z13Y9Ayn5tsmnHI6awxLl9L79toUyA/edit?usp=sharing Poster]

[[Category:Urum]] [[Category:Sp23_FinalProjects]]

User:Scasada1/Final Project

2023-05-09T15:50:55Z