Difference between revisions of "Japhug/Japhug and English"

From LING073
Jump to: navigation, search
(→‎10 sentences:)
(→‎Monolingual transducer stats)
 
(36 intermediate revisions by 3 users not shown)
Line 21: Line 21:
 
===10 sentences:===
 
===10 sentences:===
  
−
1.)
+
====#1====
−
The original sentence: kɯɕɯŋgɯ tɕe, nɤki, lɯlu ci pjɤtu, tɕendɤre βʑɯ ci pjɤtu
+
The original sentence:  
−
 
+
<pre>
 +
kɯɕɯŋgɯ tɕe, nɤki, lɯlu ci pjɤtu, tɕendɤre βʑɯ ci pjɤtu
 +
</pre>
 
The intended English translation: A long time ago, there was a cat and a mouse
 
The intended English translation: A long time ago, there was a cat and a mouse
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^kɯɕɯŋgɯ<n><ind>/long time ago<cnjadv><ind>$ ^*tɕe/*tɕe$^,<cm>/,<cm>$ ^*nɤki/*nɤki$^,<cm>/,<cm>$ ^lɯlu<n><ind>/cat<n><sg><ind>$ ^ci<det>/a<det><ind><sg>/one<num><sg>$ ^pjɤtu<cop><other>/be<vbser><other>$^,<cm>/,<cm>$ ^tɕendɤre<cnj>/then<cnjadv>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^ci<det>/a<det><ind><sg>/one<num><sg>$ ^pjɤtu<cop><other>/be<vbser><other>$^.<sent>/.<sent>$  
+
Output of jya-eng-biltrans:  
 +
<pre>
 +
^kɯɕɯŋgɯ<n><ind>/long time ago<cnjadv><ind>$ ^tɕe<cnj>//as for<pr>$^,<cm>/,<cm>$ ^nɤki<dem>/that<det><dem><sg>$^,<cm>/,<cm>$ ^lɯlu<n><ind>/cat<n><sg><ind>$ ^ci<det>/a<det><ind><sg>/one<num><sg>$ ^pjɤtu<cop><other>/be<vbser><other>$^,<cm>/,<cm>$ ^tɕendɤre<cnj>/then<cnjadv>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^ci<det>/a<det><ind><sg>/one<num><sg>$ ^pjɤtu<cop><other>/be<vbser><other>$^.<sent>/.<sent>$
 +
</pre>
  
−
The output of your xyz-abc (full translation) mode: #long time ago , that, #cat a #be, then #mouse a #be
+
Output of jya-eng:  
 +
<pre>
 +
#long time ago , that, #cat a #be, then #mouse a #be
 +
</pre>
  
  
−
 
+
====#2====
−
2.)
+
The original sentence:  
−
The original sentence: lɯlu nɯ kɯ nɤkinɯ, βʑɯ ri tundze ɕti
+
<pre>
 +
lɯlu nɯ kɯ nɤkinɯ, βʑɯ ri tundze ɕti
 +
</pre>
  
 
The intended English translation: The cat ate the mouse,
 
The intended English translation: The cat ate the mouse,
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^lɯlu<n><ind>/cat<n><sg><ind>$ ^*nɯ/*nɯ$ ^kɯ<post>/with<pr>$ ^*nɤkinɯ/*nɤkinɯ$^,<cm>/,<cm>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^ri<post>/at<pr>$ ^*tundze/*tundze$ ^ɕti<v><iv>/be<vbser>$^.<sent>/.<sent>$
+
Output of jya-eng-biltrans:
 +
<pre>
 +
^lɯlu<n><ind>/cat<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^nɤkinɯ<fill>/um<ij>$^,<cm>/,<cm>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^ri<post>/at<pr>$ ^ndze<v><tv><ipfv>/eat<vblex><ipfv>$ ^ɕti<v><iv>/be<vbser>$^.<sent>/.<sent>$
 +
</pre>
  
−
The output of your xyz-abc (full translation) mode: #cat that with um, #mouse at #eat #be
+
Output of jya-eng:
 +
<pre>
 +
#cat that with um, #mouse at #eat #be
 +
</pre>
  
 +
====#3====
 +
The original sentence:
  
−
3.)
+
<pre>
−
The original sentence: βʑɯ chondɤre nɤki, kumpɣɤtɕɯ nɯ kɯ βzaŋsa tonɯpandʑi,
+
βʑɯ chondɤre nɤki, kumpɣɤtɕɯ nɯ kɯ βzaŋsa tonɯpandʑi,
 +
</pre>
  
 
The intended English translation: The mouse made friends with the sparrow,
 
The intended English translation: The mouse made friends with the sparrow,
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^chondɤre<com>/with<pr>$ ^nɤki<dem>/that<det><dem><sg>$^,<cm>/,<cm>$ ^kumpɣɤtɕɯ<n><ind>/sparrow<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^βzaŋsa<n><ind>/friend<n><sg><ind>$ ^tonɯpandʑi<adv>/made<vblex><pp>$^,<cm>/,<cm>$^.<sent>/.<sent>$
+
Output of jya-eng-biltrans:  
 +
<pre>
 +
^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^chondɤre<com>/with<pr>$ ^nɤki<dem>/that<det><dem><sg>$^,<cm>/,<cm>$ ^kumpɣɤtɕɯ<n><ind>/sparrow<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^βzaŋsa<n><ind>/friend<n><sg><ind>$ ^tonɯpa<v><tv><du>/made<vblex><pp><du>$^,<cm>/,<cm>$^.<sent>/.<sent>$
 +
</pre>
  
−
The output of your xyz-abc (full translation) mode: #mouse with that, #sparrow that with #friend #made,
+
Output of jya-eng:
 +
<pre>
 +
#mouse with that, #sparrow that with #friend #made,
 +
</pre>
  
 +
====#4====
 +
The original sentence:
 +
<pre>
 +
tɕe βzaŋsa tonɯpandʑi tɕe, wuma pjɤkɤmɯmindʑici
 +
</pre>
  
−
4.)
+
The intended English translation: The two of them become friends and got along well
−
The original sentence: tɕe βzaŋsa tonɯpandʑi tɕe, wuma pjɤkɤmɯmindʑici
 
  
−
The intended English translation: The two of them become friends and got along well
+
Output of jya-eng-biltrans:
 +
<pre>
 +
^tɕe<cnj>//as for<pr>$ ^βzaŋsa<n><ind>/friend<n><sg><ind>$ ^tonɯpa<v><tv><du>/made<vblex><pp><du>$ ^tɕe<cnj>//as for<pr>$^,<cm>/,<cm>$ ^wuma<adv>/very<adv>$ ^ɤmɯmi<v><iv><pii><du>/get along well<vblex><pii><du>$^.<sent>/.<sent>$</pre>
 +
Output of jya-eng:
 +
<pre>
 +
#friend #made , very #get along well
 +
</pre>
 +
 
 +
====#5==== 
 +
The original sentence:
 +
<pre>
 +
maka cici ŋotɕu jɤarɯrindʑi ʑo tɯtɯrca ntsɯ juɕendʑi pjɤŋu
 +
</pre>
 +
The intended English translation: they always got together
 +
 
 +
Output of jya-eng-biltrans:
 +
<pre>
 +
^maka<adv>/at all<adv>$ ^cici<adv>/sometimes<adv>$ ^ŋotɕu<prn>/wherever<adv>$ ^*jɤarɯrindʑi/*jɤarɯrindʑi$ ^*ʑo/*ʑo$ ^*tɯtɯrca/*tɯtɯrca$ ^ntsɯ<adv>/always<adv>$ ^*juɕendʑi/*juɕendʑi$ ^ŋu<v><iv><pii>/be<vbser><pii>$^.<sent>/.<sent>$
 +
</pre>
 +
Output of jya-eng:
 +
<pre>
 +
at all sometimes wherever #each  together always #go #be
 +
</pre>
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode:
+
====#6====
 +
The original sentence:  
 +
<pre>
 +
kɤndza ci tu nɤ pjɯnɯkrondʑi
 +
</pre>
 +
The intended English translation: There was food to eat, they shared,
  
−
The output of your xyz-abc (full translation) mode: #friend #made , very *pjɤkɤmɯmindʑici
+
Output of jya-eng-biltrans:
 +
<pre>
 +
^kɤndza<n><ind>/food<n><sg><ind>$ ^ci<det>/a<det><ind><sg>/one<num><sg>$ ^tu<v><iv>/be<vbser>/exist<vblex>$ ^*nɤ/*nɤ$ ^*pjɯnɯkrondʑi/*pjɯnɯkrondʑi$^.<sent>/.<sent>$
 +
</pre>
 +
Output of jya-eng:
 +
<pre>
 +
#food a #be at #share
 +
</pre>
  
 +
====#7====
 +
The original sentence:
 +
<pre>
 +
nɯ mɤʑɯ kɯmtɕhɯ kɯfse tɤtu nɤ, tɯɣrɤz ɲɯnɤɣrondʑi nɯ ra pjɤŋgrɤl
 +
</pre>
  
−
5.)
+
The intended English translation: Also, when there was something to play with, they played together
−
The original sentence: maka cici ŋotɕu jɤarɯrindʑi ʑo tɯtɯrca ntsɯ juɕendʑi pjɤŋu
 
  
−
The intended English translation: they always got together
+
Output of jya-eng-biltrans:  
 +
<pre>
 +
^nɯ<dem>/that<det><dem><sg>$ ^mɤʑɯ<adv>/again<adv>$ ^kɯmtɕhɯ<n><ind>/toy<n><sg><ind>$ ^fse<v><tv><prc><sbj><ind>/resemble<vblex><prc><sbj><ind>$ ^tu<v><iv><aor>/be<vbser><aor>/exist<vblex><aor>$ ^nɤ<post>/at<pr>$^,<cm>/,<cm>$ ^tɯɣrɤz<adv>/together<adv>$ ^nɤɣro<v><tv><ipfv><du>/play<vblex><inf><ipfv><du>$ ^nɯ<dem>/that<det><dem><sg>$ ^ra<v><iv>/should<vbmod>$ ^ŋgrɤl<v><iv><pii>/usually like this<iv><pii>$^.<sent>/.<sent>$
 +
</pre>
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^*maka/*maka$ ^*cici/*cici$ ^ŋotɕu<prn>/wherever<prn>$ ^*jɤarɯrindʑi/*jɤarɯrindʑi$ ^*ʑo/*ʑo$ ^*tɯtɯrca/*tɯtɯrca$ ^ntsɯ<adv>/always<adv>$ ^*juɕendʑi/*juɕendʑi$ ^ŋu<v><iv><pii>/be<vbser><pii>$^.<sent>/.<sent>$
+
Output of jya-eng:
 +
<pre>
 +
that again #toy #resemble #be at, together #play that #should #usually like this
 +
</pre>
  
−
The output of your xyz-abc (full translation) mode: *maka *cici #wherever *jɤarɯrindʑi *ʑo *tɯtɯrca always *juɕendʑi #be
 
  
 +
====#8====
 +
The original sentence:
 +
<pre>
 +
tɕhi kɤcha tukɤpjɤl kowa ntsɯ tuβzundʑi pjɤŋu tɕe
 +
</pre>
  
−
6.)
+
The intended English translation: They were always trying to think of ways to evade that cat
−
The original sentence: kɤndza ci tu nɤ pjɯnɯkrondʑi
 
  
−
The intended English translation: There was food to eat, they shared,
+
Output of jya-eng-biltrans:  
 +
<pre>
 +
^tɕhi<prn>/what<prn>$ ^cha<v><iv><prc><obj><ind>/able<adj><prc><obj><ind>/can<vbmod><prc><obj><ind>$ ^pjɤl<v><tv><inf><ipfv>/evade<vblex><inf><ipfv>$ ^kowa<n><ind>/method<n><sg><ind>$ ^ntsɯ<adv>/always<adv>$ ^βzu<v><tv><ipfv><du>/make<vblex><ipfv><du>$ ^ŋu<v><iv><pii>/be<vbser><pii>$ ^tɕe<cnj>//as for<pr>$^.<sent>/.<sent>$
 +
</pre>
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^kɤndza<n><ind>/food<n><ind>$ ^*ci/*ci$ ^tu<v><iv>/be<vbser>/exist<vblex>$ ^*nɤ/*nɤ$ ^*pjɯnɯkrondʑi/*pjɯnɯkrondʑi$^.<sent>/.<sent>$
+
Output of jya-eng:
 +
<pre>
 +
#what #able #evade #method always #make #be  
 +
</pre>
  
−
The output of your xyz-abc (full translation) mode: #food *ci #be *nɤ *pjɯnɯkrondʑi
+
====#9====
 +
The original sentence:  
 +
<pre>
 +
lɯlu nɯ kɯ βʑɯ ɯɕa tɕi pjɤrga, pɣɤtɕɯ ɯɕa tɕi pjɤrga ɕti tɕe.
 +
</pre>
  
 +
The intended English translation: The cat likes eating rat meat, and also likes sparrow meat.
  
−
7.)
+
Output of jya-eng-biltrans:
−
The original sentence: nɯ mɤʑɯ kɯmtɕhɯ kɯfse tɤtu nɤ, tɯɣrɤz ɲɯnɤɣrondʑi nɯ ra pjɤŋgrɤl
+
<pre>
 +
^lɯlu<n><ind>/cat<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^ɕa<n><px3sg>/meat<n><sg><px3sg>$ ^tɕi<adv>/also<adv>$ ^rga<v><iv><pii>/like<vblex><pii>$^,<cm>/,<cm>$ ^pɣɤtɕɯ<n><ind>/bird<n><sg><ind>$ ^ɕa<n><px3sg>/meat<n><sg><px3sg>$ ^tɕi<adv>/also<adv>$ ^rga<v><iv><pii>/like<vblex><pii>$ ^ɕti<v><iv>/be<vbser>$ ^tɕe<cnj>//as for<pr>$^.<sent>/.<sent>$^.<sent>/.<sent>$
 +
</pre>
  
−
The intended English translation: Also, when there was something to play with, they played together
+
Output of jya-eng:
 +
<pre>
 +
#cat that with #mouse #meat also #like, #bird #meat also #like #be .
 +
</pre>
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^*nɯ/*nɯ$ ^mɤʑɯ<adv>/again<adv>$ ^kɯmtɕhɯ<n><ind>/toy<n><ind>$ ^fse<v><tv><prc><sbj><ind>/resemble<vblex><prc><sbj><ind>$ ^*tɤtu/*tɤtu$ ^*nɤ/*nɤ$^,<cm>/,<cm>$ ^*tɯɣrɤz/*tɯɣrɤz$ ^*ɲɯnɤɣrondʑi/*ɲɯnɤɣrondʑi$ ^*nɯ/*nɯ$ ^*ra/*ra$ ^ŋgrɤl<v><iv><pii>/usually like this<iv><pii>$^.<sent>/.<sent>$
 
  
−
The output of your xyz-abc (full translation) mode: *nɯ again #toy #resemble *tɤtu *nɤ, *tɯɣrɤz *ɲɯnɤɣrondʑi *nɯ *ra #usually like this
+
====#10====
 +
The original sentence:  
 +
<pre>
 +
tɕe nɯ maʁ nɤ tɕe, βʑɯ nɯ kɯ wuma kɤrɤfsoʁ pjɤcha tɕe
 +
</pre>
  
 +
The intended English translation: since then, that mouse was very good at accumulating food
  
−
8.)
+
Output of jya-eng-biltrans:
−
The original sentence: tɕhi kɤcha tukɤpjɤl kowa ntsɯ tuβzundʑi pjɤŋu tɕe
+
<pre>
 +
^tɕe<cnj>//as for<pr>$ ^nɯ maʁ nɤ<cnj>/or<cnjcoo>$ ^tɕe<cnj>//as for<pr>$^,<cm>/,<cm>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^wuma<adv>/very<adv>$ ^rɤfsoʁ<v><iv><prc><obj><ind>/forage<vblex><prc><obj><ind>$ ^cha<v><iv><pii>/able<adj><pii>/can<vbmod><pii>$ ^tɕe<cnj>//as for<pr>$^.<sent>/.<sent>$
 +
</pre>
  
−
The intended English translation: They were always trying to think of ways to evade that cat
+
Output of jya-eng:
 +
<pre>
 +
or , #mouse that with very #forage #able
 +
</pre>
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^tɕhi<prn>/what<prn>$ ^cha<v><iv><prc><obj><ind>/able<adj><prc><obj><ind>/can<vbmod><prc><obj><ind>$ ^*tukɤpjɤl/*tukɤpjɤl$ ^kowa<n><ind>/method<n><ind>$ ^ntsɯ<adv>/always<adv>$ ^*tuβzundʑi/*tuβzundʑi$ ^ŋu<v><iv><pii>/be<vbser><pii>$ ^*tɕe/*tɕe$^.<sent>/.<sent>$
+
===Additions ===
 +
These are some additions we made to our bilingual transducer beyond the labs.
  
−
The output of your xyz-abc (full translation) mode: #what #able *tukɤpjɤl #method always *tuβzundʑi #be *tɕe
+
* 5000+ lemmas to our Japhug transducer
  
 +
* 5 new features covered in our Japhug morphological analyzer, including: imperfective prefixes, the peg circumfix, proximal/distal demonstratives, redesigned verb patterns to match the slot-based prefix system in Japhug, number suffixes on verbs.
  
−
9.)
+
* 2 new structural transfer rules: NP rule to combine nouns and possessed nouns; DP rule to place determiner before noun, and to take ergative case from postposition; vP rule to combine adverbs and verbs. These were made in the process of meeting our original goal, to match the SVO word order in English (Japhug is usually SOV or OSV)
−
The original sentence: lɯlu nɯ kɯ βʑɯ ɯɕa tɕi pjɤrga, pɣɤtɕɯ ɯɕa tɕi pjɤrga ɕti tɕe.
 
  
−
The intended English translation: The cat likes eating rat meat, and also likes sparrow meat.
+
(details listed for our own reference)
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^lɯlu<n><ind>/cat<n><ind>$ ^*nɯ/*nɯ$ ^kɯ<post>/with<pr>$ ^βʑɯ<n><ind>/mouse<n><ind>$ ^ɕa<n><px3sg>/meat<n><px3sg>$ ^tɕi<adv>/also<adv>$ ^rga<v><iv><pii>/like<vblex><pii>$^,<cm>/,<cm>$ ^pɣɤtɕɯ<n><ind>/bird<n><ind>$ ^ɕa<n><px3sg>/meat<n><px3sg>$ ^tɕi<adv>/also<adv>$ ^rga<v><iv><pii>/like<vblex><pii>$ ^ɕti<v><iv>/be<vbser>$ ^*tɕe/*tɕe$^.<sent>/.<sent>$^.<sent>/.<sent>$
+
===Final evaluation===
  
−
The output of your xyz-abc (full translation) mode: #cat *nɯ with #mouse #meat also #like, #bird #meat also #like #be *tɕe.
+
====Monolingual transducer stats====
 +
*Updated precision and recall against the (updated) eval.test and eval.gold files:
 +
<pre>
 +
$ precisionRecall jya.eval.test.new.txt jya.eval.gold.txt
 +
Totals: 79 forms, 221 tp, 155 fp, 0 tn, 87 fn
 +
Precision: 58.77660%
 +
Recall: 71.75325%
 +
</pre>
 +
*Coverage over the large corpus:
 +
<pre>
 +
TOP UNKNOWN WORDS:
 +
    16 ^ɣɯ/*ɣɯ$
 +
    14 ^lɤβzaŋ/*lɤβzaŋ$
 +
    11 ^cho/*cho$
 +
    10 ^qartshɤs/*qartshɤs$
 +
      8 ^rɟɤlpu/*rɟɤlpu$
 +
      8 ^rcanɯ/*rcanɯ$
 +
      8 ^qhe/*qhe$
 +
      8 ^je/*je$
 +
      8 ^jamɤn/*jamɤn$
 +
      7 ^tɕɤn/*tɕɤn$
 +
      6 ^nɯmbala/*nɯmbala$
 +
      6 ^ndɤ/*ndɤ$
 +
      6 ^lɤari/*lɤari$
 +
      6 ^kɯki/*kɯki$
 +
      5 ^wo/*wo$
 +
      5 ^tɯɣjɤn/*tɯɣjɤn$
 +
      5 ^lɤɣe/*lɤɣe$
 +
      5 ^kɯsthɯci/*kɯsthɯci$
 +
      5 ^ciŋglɯɣ/*ciŋglɯɣ$
 +
      4 ^ɲɯsɯsɤm/*ɲɯsɯsɤm$
 +
coverage: 4463 / 5019 (~0.88922096035066746364)
 +
remaining unknown forms: 556
 +
ambiguity: 9939 / 5019 (~1.98027495517035265989)
 +
jya.corpus Mon May 8 02:13:49 PM EDT 2023      4463/5019      ~0.88922096035066746364  \|      9939/5019      ~1.98027495517035265989
 +
</pre>
 +
*The number of words in the large corpus: 3999
 +
*The number of stems in the transducer: 6646
  
 +
====MT in jya-eng stats====
 +
*WER and PER over longer corpus:
  
−
10.)  
+
<pre>
−
The original sentence: tɕe nɯ maʁ nɤ tɕe, βʑɯ nɯ kɯ wuma kɤrɤfsoʁ pjɤcha tɕe
+
Statistics about input files
 +
-------------------------------------------------------
 +
Number of words in reference: 3013
 +
Number of words in test: 2450
 +
Number of unknown words (marked with a star) in test: 996
 +
Percentage of unknown words: 40.65 %
  
−
The intended English translation: since then, that mouse was very good at accumulating food
+
Results when removing unknown-word marks (stars)
 +
-------------------------------------------------------
 +
Edit distance: 3116
 +
Word error rate (WER): 103.42 %
 +
Number of position-independent correct words: 85
 +
Position-independent word error rate (PER): 97.18 %
 +
</pre>
 +
*The proportion of stems translated correctly in the longer corpus: 100% - 103.42% = -3.42%
 +
*Trimmed coverage over longer corpora: ~0.68006279434850863422
 +
*Trimmed coverage over large corpora: ~0.63545647558386411890
 +
*The number of tokens in longer corpora: 3185
 +
*The number of tokens in large corpora: 4710
  
−
The output of your xyz-abc-biltrans (lexical transfer) mode: ^*tɕe/*tɕe$ ^*nɯ/*nɯ$ ^*maʁ/*maʁ$ ^*nɤ/*nɤ$ ^*tɕe/*tɕe$^,<cm>/,<cm>$ ^βʑɯ<n><ind>/mouse<n><ind>$ ^*nɯ/*nɯ$ ^kɯ<post>/with<pr>$ ^wuma<adv>/very<adv>$ ^rɤfsoʁ<v><iv><prc><obj><ind>/forage<vblex><prc><obj><ind>$ ^cha<v><iv><pii>/able<adj><pii>/can<vbmod><pii>$ ^*tɕe/*tɕe$^.<sent>/.<sent>$
 
  
−
The output of your xyz-abc (full translation) mode: *tɕe *nɯ *maʁ *nɤ *tɕe, #mouse *nɯ with very #forage #able *tɕe
 
  
−
 
 
[[Category:Sp23_TranslationPairs ]] [[Category: Japhug]]
 
[[Category:Sp23_TranslationPairs ]] [[Category: Japhug]]

Latest revision as of 02:48, 10 May 2023

Resources for machine translation between Japhug and English

Developed resources

Japhug-English corpus


Lexical selection

External resources

Apertium language pair for translating between Jiarong and English

Monolingual Japhug transducer

Jya → Eng evaluation

coverage (number of forms analysed) of Japhug transducer on jya.sentences.txt: 76.98 %

coverage of bilingual transducer (jya-eng.automorf.bin) on the same file: 28.61 %

10 sentences:

#1

The original sentence: 

kɯɕɯŋgɯ tɕe, nɤki, lɯlu ci pjɤtu, tɕendɤre βʑɯ ci pjɤtu

The intended English translation: A long time ago, there was a cat and a mouse

Output of jya-eng-biltrans: 

^kɯɕɯŋgɯ<n><ind>/long time ago<cnjadv><ind>$ ^tɕe<cnj>//as for<pr>$^,<cm>/,<cm>$ ^nɤki<dem>/that<det><dem><sg>$^,<cm>/,<cm>$ ^lɯlu<n><ind>/cat<n><sg><ind>$ ^ci<det>/a<det><ind><sg>/one<num><sg>$ ^pjɤtu<cop><other>/be<vbser><other>$^,<cm>/,<cm>$ ^tɕendɤre<cnj>/then<cnjadv>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^ci<det>/a<det><ind><sg>/one<num><sg>$ ^pjɤtu<cop><other>/be<vbser><other>$^.<sent>/.<sent>$

Output of jya-eng: 

#long time ago , that, #cat a #be, then #mouse a #be


#2

The original sentence: 

lɯlu nɯ kɯ nɤkinɯ, βʑɯ ri tundze ɕti

The intended English translation: The cat ate the mouse,

Output of jya-eng-biltrans: 

^lɯlu<n><ind>/cat<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^nɤkinɯ<fill>/um<ij>$^,<cm>/,<cm>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^ri<post>/at<pr>$ ^ndze<v><tv><ipfv>/eat<vblex><ipfv>$ ^ɕti<v><iv>/be<vbser>$^.<sent>/.<sent>$

Output of jya-eng: 

#cat that with um, #mouse at #eat #be

#3

The original sentence: 

βʑɯ chondɤre nɤki, kumpɣɤtɕɯ nɯ kɯ βzaŋsa tonɯpandʑi,

The intended English translation: The mouse made friends with the sparrow,

Output of jya-eng-biltrans: 

^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^chondɤre<com>/with<pr>$ ^nɤki<dem>/that<det><dem><sg>$^,<cm>/,<cm>$ ^kumpɣɤtɕɯ<n><ind>/sparrow<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^βzaŋsa<n><ind>/friend<n><sg><ind>$ ^tonɯpa<v><tv><du>/made<vblex><pp><du>$^,<cm>/,<cm>$^.<sent>/.<sent>$

Output of jya-eng: 

#mouse with that, #sparrow that with #friend #made,

#4

The original sentence: 

tɕe βzaŋsa tonɯpandʑi tɕe, wuma pjɤkɤmɯmindʑici

The intended English translation: The two of them become friends and got along well

Output of jya-eng-biltrans: 

^tɕe<cnj>//as for<pr>$ ^βzaŋsa<n><ind>/friend<n><sg><ind>$ ^tonɯpa<v><tv><du>/made<vblex><pp><du>$ ^tɕe<cnj>//as for<pr>$^,<cm>/,<cm>$ ^wuma<adv>/very<adv>$ ^ɤmɯmi<v><iv><pii><du>/get along well<vblex><pii><du>$^.<sent>/.<sent>$

Output of jya-eng: 

#friend #made , very #get along well

#5

The original sentence: 

maka cici ŋotɕu jɤarɯrindʑi ʑo tɯtɯrca ntsɯ juɕendʑi pjɤŋu

The intended English translation: they always got together

Output of jya-eng-biltrans: 

^maka<adv>/at all<adv>$ ^cici<adv>/sometimes<adv>$ ^ŋotɕu<prn>/wherever<adv>$ ^*jɤarɯrindʑi/*jɤarɯrindʑi$ ^*ʑo/*ʑo$ ^*tɯtɯrca/*tɯtɯrca$ ^ntsɯ<adv>/always<adv>$ ^*juɕendʑi/*juɕendʑi$ ^ŋu<v><iv><pii>/be<vbser><pii>$^.<sent>/.<sent>$

Output of jya-eng: 

at all sometimes wherever #each  together always #go #be

#6

The original sentence: 

kɤndza ci tu nɤ pjɯnɯkrondʑi

The intended English translation: There was food to eat, they shared,

Output of jya-eng-biltrans: 

^kɤndza<n><ind>/food<n><sg><ind>$ ^ci<det>/a<det><ind><sg>/one<num><sg>$ ^tu<v><iv>/be<vbser>/exist<vblex>$ ^*nɤ/*nɤ$ ^*pjɯnɯkrondʑi/*pjɯnɯkrondʑi$^.<sent>/.<sent>$

Output of jya-eng: 

#food a #be at #share

#7

The original sentence: 

nɯ mɤʑɯ kɯmtɕhɯ kɯfse tɤtu nɤ, tɯɣrɤz ɲɯnɤɣrondʑi nɯ ra pjɤŋgrɤl

The intended English translation: Also, when there was something to play with, they played together

Output of jya-eng-biltrans: 

^nɯ<dem>/that<det><dem><sg>$ ^mɤʑɯ<adv>/again<adv>$ ^kɯmtɕhɯ<n><ind>/toy<n><sg><ind>$ ^fse<v><tv><prc><sbj><ind>/resemble<vblex><prc><sbj><ind>$ ^tu<v><iv><aor>/be<vbser><aor>/exist<vblex><aor>$ ^nɤ<post>/at<pr>$^,<cm>/,<cm>$ ^tɯɣrɤz<adv>/together<adv>$ ^nɤɣro<v><tv><ipfv><du>/play<vblex><inf><ipfv><du>$ ^nɯ<dem>/that<det><dem><sg>$ ^ra<v><iv>/should<vbmod>$ ^ŋgrɤl<v><iv><pii>/usually like this<iv><pii>$^.<sent>/.<sent>$

Output of jya-eng: 

that again #toy #resemble #be at, together #play that #should #usually like this


#8

The original sentence: 

tɕhi kɤcha tukɤpjɤl kowa ntsɯ tuβzundʑi pjɤŋu tɕe

The intended English translation: They were always trying to think of ways to evade that cat

Output of jya-eng-biltrans: 

^tɕhi<prn>/what<prn>$ ^cha<v><iv><prc><obj><ind>/able<adj><prc><obj><ind>/can<vbmod><prc><obj><ind>$ ^pjɤl<v><tv><inf><ipfv>/evade<vblex><inf><ipfv>$ ^kowa<n><ind>/method<n><sg><ind>$ ^ntsɯ<adv>/always<adv>$ ^βzu<v><tv><ipfv><du>/make<vblex><ipfv><du>$ ^ŋu<v><iv><pii>/be<vbser><pii>$ ^tɕe<cnj>//as for<pr>$^.<sent>/.<sent>$

Output of jya-eng: 

#what #able #evade #method always #make #be

#9

The original sentence: 

lɯlu nɯ kɯ βʑɯ ɯɕa tɕi pjɤrga, pɣɤtɕɯ ɯɕa tɕi pjɤrga ɕti tɕe.

The intended English translation: The cat likes eating rat meat, and also likes sparrow meat.

Output of jya-eng-biltrans: 

^lɯlu<n><ind>/cat<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^ɕa<n><px3sg>/meat<n><sg><px3sg>$ ^tɕi<adv>/also<adv>$ ^rga<v><iv><pii>/like<vblex><pii>$^,<cm>/,<cm>$ ^pɣɤtɕɯ<n><ind>/bird<n><sg><ind>$ ^ɕa<n><px3sg>/meat<n><sg><px3sg>$ ^tɕi<adv>/also<adv>$ ^rga<v><iv><pii>/like<vblex><pii>$ ^ɕti<v><iv>/be<vbser>$ ^tɕe<cnj>//as for<pr>$^.<sent>/.<sent>$^.<sent>/.<sent>$

Output of jya-eng: 

#cat that with #mouse #meat also #like, #bird #meat also #like #be .


#10

The original sentence: 

tɕe nɯ maʁ nɤ tɕe, βʑɯ nɯ kɯ wuma kɤrɤfsoʁ pjɤcha tɕe

The intended English translation: since then, that mouse was very good at accumulating food

Output of jya-eng-biltrans: 

^tɕe<cnj>//as for<pr>$ ^nɯ maʁ nɤ<cnj>/or<cnjcoo>$ ^tɕe<cnj>//as for<pr>$^,<cm>/,<cm>$ ^βʑɯ<n><ind>/mouse<n><sg><ind>$ ^nɯ<dem>/that<det><dem><sg>$ ^kɯ<post>/with<pr>$ ^wuma<adv>/very<adv>$ ^rɤfsoʁ<v><iv><prc><obj><ind>/forage<vblex><prc><obj><ind>$ ^cha<v><iv><pii>/able<adj><pii>/can<vbmod><pii>$ ^tɕe<cnj>//as for<pr>$^.<sent>/.<sent>$

Output of jya-eng: 

or , #mouse that with very #forage #able

Additions

These are some additions we made to our bilingual transducer beyond the labs.

  • 5000+ lemmas to our Japhug transducer
  • 5 new features covered in our Japhug morphological analyzer, including: imperfective prefixes, the peg circumfix, proximal/distal demonstratives, redesigned verb patterns to match the slot-based prefix system in Japhug, number suffixes on verbs.
  • 2 new structural transfer rules: NP rule to combine nouns and possessed nouns; DP rule to place determiner before noun, and to take ergative case from postposition; vP rule to combine adverbs and verbs. These were made in the process of meeting our original goal, to match the SVO word order in English (Japhug is usually SOV or OSV)

(details listed for our own reference)

Final evaluation

Monolingual transducer stats

  • Updated precision and recall against the (updated) eval.test and eval.gold files: 
$ precisionRecall jya.eval.test.new.txt jya.eval.gold.txt 
Totals: 79 forms, 221 tp, 155 fp, 0 tn, 87 fn
Precision: 58.77660%
Recall: 71.75325%
  • Coverage over the large corpus:
TOP UNKNOWN WORDS:
     16 ^ɣɯ/*ɣɯ$
     14 ^lɤβzaŋ/*lɤβzaŋ$
     11 ^cho/*cho$
     10 ^qartshɤs/*qartshɤs$
      8 ^rɟɤlpu/*rɟɤlpu$
      8 ^rcanɯ/*rcanɯ$
      8 ^qhe/*qhe$
      8 ^je/*je$
      8 ^jamɤn/*jamɤn$
      7 ^tɕɤn/*tɕɤn$
      6 ^nɯmbala/*nɯmbala$
      6 ^ndɤ/*ndɤ$
      6 ^lɤari/*lɤari$
      6 ^kɯki/*kɯki$
      5 ^wo/*wo$
      5 ^tɯɣjɤn/*tɯɣjɤn$
      5 ^lɤɣe/*lɤɣe$
      5 ^kɯsthɯci/*kɯsthɯci$
      5 ^ciŋglɯɣ/*ciŋglɯɣ$
      4 ^ɲɯsɯsɤm/*ɲɯsɯsɤm$
coverage: 4463 / 5019 (~0.88922096035066746364)
remaining unknown forms: 556
ambiguity: 9939 / 5019 (~1.98027495517035265989)
jya.corpus Mon May 8 02:13:49 PM EDT 2023       4463/5019       ~0.88922096035066746364  \|      9939/5019       ~1.98027495517035265989
  • The number of words in the large corpus: 3999
  • The number of stems in the transducer: 6646

MT in jya-eng stats

  • WER and PER over longer corpus:
Statistics about input files
-------------------------------------------------------
Number of words in reference: 3013
Number of words in test: 2450
Number of unknown words (marked with a star) in test: 996
Percentage of unknown words: 40.65 %

Results when removing unknown-word marks (stars)
-------------------------------------------------------
Edit distance: 3116
Word error rate (WER): 103.42 %
Number of position-independent correct words: 85
Position-independent word error rate (PER): 97.18 %
  • The proportion of stems translated correctly in the longer corpus: 100% - 103.42% = -3.42%
  • Trimmed coverage over longer corpora: ~0.68006279434850863422
  • Trimmed coverage over large corpora: ~0.63545647558386411890
  • The number of tokens in longer corpora: 3185
  • The number of tokens in large corpora: 4710
Retrieved from "http://wikis.swarthmore.edu/ling073/index.php?title=Japhug/Japhug_and_English&oldid=22489"