Difference between revisions of "Uyghur and English"
From LING073
(→Lexical Transfer) |
(→Coverage) |
||
| (18 intermediate revisions by 2 users not shown) | |||
| Line 1: | Line 1: | ||
| − | This page contains the resources for the Uyghur and English. | + | This page contains the resources for the [https://wikis.swarthmore.edu/ling073/Uyghur Uyghur] and [https://wikis.swarthmore.edu/ling073/English English]. |
| − | ==Developed Uyghur Resources== | + | ===Developed Uyghur Resources=== |
*[https://wikis.swarthmore.edu/ling073/Uyghur Main Wiki Page] | *[https://wikis.swarthmore.edu/ling073/Uyghur Main Wiki Page] | ||
*[https://github.swarthmore.edu/Ling073-sp25/ling073-uig-corpus Uyghur corpus repository] | *[https://github.swarthmore.edu/Ling073-sp25/ling073-uig-corpus Uyghur corpus repository] | ||
| Line 7: | Line 7: | ||
*[https://wikis.swarthmore.edu/ling073/Uyghur/Grammar Uyghur Grammar ] | *[https://wikis.swarthmore.edu/ling073/Uyghur/Grammar Uyghur Grammar ] | ||
*[https://wikis.swarthmore.edu/ling073/Uyghur/Transducer Uyghur Transducer and Evaluation] | *[https://wikis.swarthmore.edu/ling073/Uyghur/Transducer Uyghur Transducer and Evaluation] | ||
| + | *[https://wikis.swarthmore.edu/ling073/Uyghur_and_English/Contrastive_Grammar Uyghur and English Contrastive Grammar] | ||
| − | ==External Resources | + | ===External Resources=== |
| − | *[https:// | + | *[https://github.com/apertium/apertium-eng Apertium English Repository] |
| − | *[https://github. | + | *[https://github.swarthmore.edu/Ling073-sp25/ling073-uig Uyghur Transducer] |
| + | *[https://github.swarthmore.edu/Ling073-sp25/ling073-uig-eng Uyghur-English Bilingual Dictionary] | ||
| − | === | + | ===xyz → abc Evaluation=== |
| − | + | ====Uyghur Coverage==== | |
| − | * | + | *coverage: 206 / 215 (~0.95813953488372093023) |
| − | * | + | *remaining unknown forms: 9 |
| − | * | + | *ambiguity: 254 / 215 (~1.18139534883720930233) |
| + | *uig.sentences.txt Wed May 7 06:01:48 PM EDT 2025 206/215 ~0.95813953488372093023 \| 254/215 ~1.18139534883720930233 | ||
| + | ====Bilingual Coverage==== | ||
| + | *coverage: 165 / 215 (~0.76744186046511627907) | ||
| + | *remaining unknown forms: 50 | ||
| + | *ambiguity: 240 / 215 (~1.11627906976744186047) | ||
| + | *uig.sentences.txt Wed May 7 06:06:05 PM EDT 2025 165/215 ~0.76744186046511627907 \| 240/215 ~1.11627906976744186047 | ||
| + | |||
| + | ====10 Original Sentences==== | ||
| + | خۇدا ھەممىدىن بۇرۇن پۈتۈن ئالەمنى ياراتتى. | ||
| + | ئۇ چاغدا يەر شارى شەكىلسىز ۋە قۇپقۇرۇق ھالەتتە بولۇپ سۇلارنى قاراڭغۇلۇق قاپلىغانىدى. | ||
| + | خۇدانىڭ روھى سۇ ئۈستىدە لەرزان پەرۋاز قىلىپ يۈرەتتى. | ||
| + | خۇدا: «يورۇقلۇق بولسۇن!» دېدى يورۇقلۇق پەيدا بولدى. | ||
| + | خۇدا يورۇقلۇقنىڭ ياخشى ئىكەنلىكىنى كۆردى. | ||
| + | شۇنىڭ بىلەن ئۇ يورۇقلۇقنى قاراڭغۇلۇقتىن ئايرىدى ۋە يورۇقلۇقنى «كۈندۈز» قاراڭغۇلۇقنى «كېچە» دەپ ئاتىدى. | ||
| + | كېچە ۋە كۈندۈز ئۆتتى بۇ تۇنجى كۈن بولدى. | ||
| + | ئاندىن خۇدا: «سۇلارنى ئىككىگە ئايرىش ئۈچۈن بىر پاسىل بولسۇن!» دېدى. | ||
| + | خۇدانىڭ ئەمرى بىلەن شۇنداق بولدى. | ||
| + | خۇدا بۇ پاسىلنى «ئاسمان» دەپ ئاتىدى. | ||
| + | |||
| + | ====Intended Translations==== | ||
| + | God created the entire universe first. | ||
| + | At that time, the earth was formless and empty, and darkness covered the waters. | ||
| + | The Spirit of God was hovering over the waters. | ||
| + | God said, "Let it be very bright!" and it was very bright. | ||
| + | God saw that the light was good. | ||
| + | So he separated the light from the darkness and called the light "day" and the darkness "night." | ||
| + | Night and day passed, and this was the first day. | ||
| + | Then God said, "Let there be a firmament to separate the waters!" | ||
| + | And with God's command, it was so. | ||
| + | And God called the firmament Heaven. | ||
| + | |||
| + | ====uig-eng==== | ||
| + | * #God #all #before entire #universe #create. | ||
| + | * #this #time #land #globe shapeless and #empty #state #be #water #darkness #cover. | ||
| + | * #God #spirit #water #top shaky #murmuration #do #go. | ||
| + | * #God #brightness #be #say #brightness #appearance #be. | ||
| + | * #God #brightness #good #that be #see. | ||
| + | * #that #through #this #brightness #darkness #separate and #brightness «#daytime» #darkness «#night» #call. | ||
| + | * #night and #daytime #pass #this first #day #be. | ||
| + | * then #God #water #two #separation #for #a #boundary #be #say. | ||
| + | * #God #command #through so #be. | ||
| + | * #God #this #boundary «#heaven» #call. | ||
| + | |||
| + | ====uig-eng-biltrans==== | ||
| + | *^خۇدا<n><nom>/God<n><nom>$ ^ھەممې<prn><ind><abl>/all<prn><qnt><mf><sp><abl>$ ^بۇرۇن<post><nom>/before<pr><nom>$ ^پۈتۈن<adj>/entire<adj>$ ^ئالەم<n><acc>/universe<n><acc>$ ^يارات<vblex><pas><p3>/create<vblex><pas><p3>$. | ||
| + | *^ئۇ<det>/this<det><dem>$ ^چاغ<n><loc>/time<n><loc>$ ^يەر<n><nom>/land<n><nom>$ ^شارى<n><nom>/globe<n><nom>$ ^شەكىل<adj>/shapeless<adj>$ ^ۋە<cnjcoo>/and<cnjcoo>$ ^قۇرۇق<adj><emp>/empty<adj><emp>$ ^ھالەت<n><loc>/state<n><loc>$ ^بول<cnjadv>/be<vbser>$ ^سۇ<n><pl><acc>/water<n><pl><acc>$ ^قاراڭغۇلۇق<n><nom>/darkness<n><nom>$ ^قاپلى<vblex><perf><pas><p3>/cover<vblex><perf><pas><p3>$. | ||
| + | *^خۇدا<n><gen>/God<n><gen>$ ^روھى<n><nom>/spirit<n><nom>$ ^سۇ<n><nom>/water<n><nom>$ ^ئۈستى<n><loc>/top<n><loc>$ ^لەرزان<adj>/shaky<adj><sint>$ ^پەرۋاز<n><nom>/murmuration<n><nom>$ ^قىل<cnjadv>/do<vbdo>$ ^يۈر<vblex><impf><pas><p3>/go<vblex><impf><pas><p3>$. | ||
| + | *^خۇدا<n><nom>/God<n><nom>$ ^يورۇق<n><nom>/brightness<n><nom>$ ^بول<vblex><imp><p3>/be<vbser><imp><p3>$ ^دې<vblex><pas><p3>/say<vblex><pas><p3>$ ^يورۇق<n><nom>/brightness<n><nom>$ ^پەيدا<n><nom>/appearance<n><nom>$ ^بول<vblex><pas><p3>/be<vbser><pas><p3>$. | ||
| + | *^خۇدا<n><nom>/God<n><nom>$ ^يورۇق<n><gen>/brightness<n><gen>$ ^ياخشى<adj>/good<adj>$ ^ئىكەن<cop><p3><n><acc>/that be<n><acc>$ ^كۆر<vblex><pas><p3>/see<vblex><pas><p3>$. | ||
| + | *^شۇ<prn><dem><gen>/that<n><dem><mf><gen>$ ^بىلەن<post><nom>/through<pr><nom>$ ^ئۇ<det>/this<det><dem>$ ^يورۇق<n><acc>/brightness<n><acc>$ ^قاراڭغۇلۇق<n><abl>/darkness<n><abl>$ ^ئايرى<vblex><pas><p3>/separate<vblex><pas><p3>$ ^ۋە<cnjcoo>/and<cnjcoo>$ ^يورۇق<n><acc>/brightness<n><acc>$ «^كۈندۈز<n><nom>/daytime<n><nom>$» ^قاراڭغۇلۇق<n><acc>/darkness<n><acc>$ «^كېچە<n><nom>/night<n><nom>$» ^دە<cnjadv>/$ ^ئاتى<vblex><pas><p3>/call<vblex><pas><p3>$. | ||
| + | *^كېچە<n><nom>/night<n><nom>$ ^ۋە<cnjcoo>/and<cnjcoo>$ ^كۈندۈز<n><nom>/daytime<n><nom>$ ^ئۆت<vblex><pas><p3>/pass<vblex><pas><p3>$ ^بۇ<prn><dem><nom>/this<prn><dem><nom>$ ^تۇنجى<adj>/first<adj>$ ^كۈن<n><nom>/day<n><nom>$ ^بول<vblex><pas><p3>/be<vbser><pas><p3>$. | ||
| + | *^ئاندىن<adv>/then<adv>$ ^خۇدا<n><nom>/God<n><nom>$ ^سۇ<n><pl><acc>/water<n><pl><acc>$ ^ئىككى<num><dat>/two<num><dat>$ ^ئايرى<ger>/separation<n>$ ^ئۈچۈن<post><nom>/for<pr><nom>$ ^بىر<det>/a<det>$ ^پاسىل<n><nom>/boundary<n><nom>$ ^بول<vblex><imp><p3>/be<vbser><imp><p3>$ ^دې<vblex><pas><p3>/say<vblex><pas><p3>$. | ||
| + | *^خۇدا<n><nom>/God<n><nom>$ ^بۇ<prn><dem><nom>/this<prn><dem><nom>$ ^پاسىل<n><acc>/boundary<n><acc>$ «^ئاسمان<n><nom>/heaven<n><nom>$» ^دە<cnjadv>/$ ^ئاتى<vblex><pas><p3>/call<vblex><pas><p3>$. | ||
| + | |||
| + | ===Additions=== | ||
| + | *Added around 10 structural transfer rules | ||
| + | *Added 5 new twol rules | ||
| + | *Added 5 morphological elements focusing on verb forms and participles | ||
| + | |||
| + | ====Coverage==== | ||
| + | Large Corpus: | ||
| + | coverage: 662493 / 2667733 (~0.24833557181322118818) | ||
| + | remaining unknown forms: 2005240 | ||
| + | ambiguity: 2882616 / 2667733 (~1.08054891550241347241) | ||
| + | uig.corpus Wed May 14 12:18:05 AM EDT 2025 662493/2667733 ~0.24833557181322118818 \| 2882616/2667733 ~1.08054891550241347241 | ||
| + | |||
| + | With 130 stems in the transducer. | ||
| + | |||
| + | ====uig-eng==== | ||
| + | Statistics about input files | ||
| + | ------------------------------------------------------- | ||
| + | Number of words in reference: 332 | ||
| + | Number of words in test: 332 | ||
| + | Number of unknown words (marked with a star) in test: 45 | ||
| + | Percentage of unknown words: 13.55 % | ||
| + | |||
| + | Results when removing unknown-word marks (stars) | ||
| + | ------------------------------------------------------- | ||
| + | Edit distance: 288 | ||
| + | Word error rate (WER): 86.75 % | ||
| + | Number of position-independent correct words: 94 | ||
| + | Position-independent word error rate (PER): 71.69 % | ||
[[Category:Sp25_TranslationPairs]] | [[Category:Sp25_TranslationPairs]] | ||
Latest revision as of 04:31, 14 May 2025
This page contains the resources for the Uyghur and English.
Contents
Developed Uyghur Resources
- Main Wiki Page
- Uyghur corpus repository
- Uyghur Keyboard
- Uyghur Grammar
- Uyghur Transducer and Evaluation
- Uyghur and English Contrastive Grammar
External Resources
xyz → abc Evaluation
Uyghur Coverage
- coverage: 206 / 215 (~0.95813953488372093023)
- remaining unknown forms: 9
- ambiguity: 254 / 215 (~1.18139534883720930233)
- uig.sentences.txt Wed May 7 06:01:48 PM EDT 2025 206/215 ~0.95813953488372093023 \| 254/215 ~1.18139534883720930233
Bilingual Coverage
- coverage: 165 / 215 (~0.76744186046511627907)
- remaining unknown forms: 50
- ambiguity: 240 / 215 (~1.11627906976744186047)
- uig.sentences.txt Wed May 7 06:06:05 PM EDT 2025 165/215 ~0.76744186046511627907 \| 240/215 ~1.11627906976744186047
10 Original Sentences
خۇدا ھەممىدىن بۇرۇن پۈتۈن ئالەمنى ياراتتى. ئۇ چاغدا يەر شارى شەكىلسىز ۋە قۇپقۇرۇق ھالەتتە بولۇپ سۇلارنى قاراڭغۇلۇق قاپلىغانىدى. خۇدانىڭ روھى سۇ ئۈستىدە لەرزان پەرۋاز قىلىپ يۈرەتتى. خۇدا: «يورۇقلۇق بولسۇن!» دېدى يورۇقلۇق پەيدا بولدى. خۇدا يورۇقلۇقنىڭ ياخشى ئىكەنلىكىنى كۆردى. شۇنىڭ بىلەن ئۇ يورۇقلۇقنى قاراڭغۇلۇقتىن ئايرىدى ۋە يورۇقلۇقنى «كۈندۈز» قاراڭغۇلۇقنى «كېچە» دەپ ئاتىدى. كېچە ۋە كۈندۈز ئۆتتى بۇ تۇنجى كۈن بولدى. ئاندىن خۇدا: «سۇلارنى ئىككىگە ئايرىش ئۈچۈن بىر پاسىل بولسۇن!» دېدى. خۇدانىڭ ئەمرى بىلەن شۇنداق بولدى. خۇدا بۇ پاسىلنى «ئاسمان» دەپ ئاتىدى.
Intended Translations
God created the entire universe first. At that time, the earth was formless and empty, and darkness covered the waters. The Spirit of God was hovering over the waters. God said, "Let it be very bright!" and it was very bright. God saw that the light was good. So he separated the light from the darkness and called the light "day" and the darkness "night." Night and day passed, and this was the first day. Then God said, "Let there be a firmament to separate the waters!" And with God's command, it was so. And God called the firmament Heaven.
uig-eng
- #God #all #before entire #universe #create.
- #this #time #land #globe shapeless and #empty #state #be #water #darkness #cover.
- #God #spirit #water #top shaky #murmuration #do #go.
- #God #brightness #be #say #brightness #appearance #be.
- #God #brightness #good #that be #see.
- #that #through #this #brightness #darkness #separate and #brightness «#daytime» #darkness «#night» #call.
- #night and #daytime #pass #this first #day #be.
- then #God #water #two #separation #for #a #boundary #be #say.
- #God #command #through so #be.
- #God #this #boundary «#heaven» #call.
uig-eng-biltrans
- ^خۇدا<n><nom>/God<n><nom>$ ^ھەممې<prn><ind><abl>/all<prn><qnt><mf><sp><abl>$ ^بۇرۇن<post><nom>/before<pr><nom>$ ^پۈتۈن<adj>/entire<adj>$ ^ئالەم<n><acc>/universe<n><acc>$ ^يارات<vblex><pas><p3>/create<vblex><pas><p3>$.
- ^ئۇ<det>/this<det><dem>$ ^چاغ<n><loc>/time<n><loc>$ ^يەر<n><nom>/land<n><nom>$ ^شارى<n><nom>/globe<n><nom>$ ^شەكىل<adj>/shapeless<adj>$ ^ۋە<cnjcoo>/and<cnjcoo>$ ^قۇرۇق<adj><emp>/empty<adj><emp>$ ^ھالەت<n><loc>/state<n><loc>$ ^بول<cnjadv>/be<vbser>$ ^سۇ<n><pl><acc>/water<n><pl><acc>$ ^قاراڭغۇلۇق<n><nom>/darkness<n><nom>$ ^قاپلى<vblex><perf><pas><p3>/cover<vblex><perf><pas><p3>$.
- ^خۇدا<n><gen>/God<n><gen>$ ^روھى<n><nom>/spirit<n><nom>$ ^سۇ<n><nom>/water<n><nom>$ ^ئۈستى<n><loc>/top<n><loc>$ ^لەرزان<adj>/shaky<adj><sint>$ ^پەرۋاز<n><nom>/murmuration<n><nom>$ ^قىل<cnjadv>/do<vbdo>$ ^يۈر<vblex><impf><pas><p3>/go<vblex><impf><pas><p3>$.
- ^خۇدا<n><nom>/God<n><nom>$ ^يورۇق<n><nom>/brightness<n><nom>$ ^بول<vblex><imp><p3>/be<vbser><imp><p3>$ ^دې<vblex><pas><p3>/say<vblex><pas><p3>$ ^يورۇق<n><nom>/brightness<n><nom>$ ^پەيدا<n><nom>/appearance<n><nom>$ ^بول<vblex><pas><p3>/be<vbser><pas><p3>$.
- ^خۇدا<n><nom>/God<n><nom>$ ^يورۇق<n><gen>/brightness<n><gen>$ ^ياخشى<adj>/good<adj>$ ^ئىكەن<cop><p3><n><acc>/that be<n><acc>$ ^كۆر<vblex><pas><p3>/see<vblex><pas><p3>$.
- ^شۇ<prn><dem><gen>/that<n><dem><mf><gen>$ ^بىلەن<post><nom>/through<pr><nom>$ ^ئۇ<det>/this<det><dem>$ ^يورۇق<n><acc>/brightness<n><acc>$ ^قاراڭغۇلۇق<n><abl>/darkness<n><abl>$ ^ئايرى<vblex><pas><p3>/separate<vblex><pas><p3>$ ^ۋە<cnjcoo>/and<cnjcoo>$ ^يورۇق<n><acc>/brightness<n><acc>$ «^كۈندۈز<n><nom>/daytime<n><nom>$» ^قاراڭغۇلۇق<n><acc>/darkness<n><acc>$ «^كېچە<n><nom>/night<n><nom>$» ^دە<cnjadv>/$ ^ئاتى<vblex><pas><p3>/call<vblex><pas><p3>$.
- ^كېچە<n><nom>/night<n><nom>$ ^ۋە<cnjcoo>/and<cnjcoo>$ ^كۈندۈز<n><nom>/daytime<n><nom>$ ^ئۆت<vblex><pas><p3>/pass<vblex><pas><p3>$ ^بۇ<prn><dem><nom>/this<prn><dem><nom>$ ^تۇنجى<adj>/first<adj>$ ^كۈن<n><nom>/day<n><nom>$ ^بول<vblex><pas><p3>/be<vbser><pas><p3>$.
- ^ئاندىن<adv>/then<adv>$ ^خۇدا<n><nom>/God<n><nom>$ ^سۇ<n><pl><acc>/water<n><pl><acc>$ ^ئىككى<num><dat>/two<num><dat>$ ^ئايرى<ger>/separation<n>$ ^ئۈچۈن<post><nom>/for<pr><nom>$ ^بىر<det>/a<det>$ ^پاسىل<n><nom>/boundary<n><nom>$ ^بول<vblex><imp><p3>/be<vbser><imp><p3>$ ^دې<vblex><pas><p3>/say<vblex><pas><p3>$.
- ^خۇدا<n><nom>/God<n><nom>$ ^بۇ<prn><dem><nom>/this<prn><dem><nom>$ ^پاسىل<n><acc>/boundary<n><acc>$ «^ئاسمان<n><nom>/heaven<n><nom>$» ^دە<cnjadv>/$ ^ئاتى<vblex><pas><p3>/call<vblex><pas><p3>$.
Additions
- Added around 10 structural transfer rules
- Added 5 new twol rules
- Added 5 morphological elements focusing on verb forms and participles
Coverage
Large Corpus:
coverage: 662493 / 2667733 (~0.24833557181322118818) remaining unknown forms: 2005240 ambiguity: 2882616 / 2667733 (~1.08054891550241347241) uig.corpus Wed May 14 12:18:05 AM EDT 2025 662493/2667733 ~0.24833557181322118818 \| 2882616/2667733 ~1.08054891550241347241
With 130 stems in the transducer.
uig-eng
Statistics about input files ------------------------------------------------------- Number of words in reference: 332 Number of words in test: 332 Number of unknown words (marked with a star) in test: 45 Percentage of unknown words: 13.55 %
Results when removing unknown-word marks (stars) ------------------------------------------------------- Edit distance: 288 Word error rate (WER): 86.75 % Number of position-independent correct words: 94 Position-independent word error rate (PER): 71.69 %
