Difference between revisions of "Mixe and English"

From LING073
Jump to: navigation, search
(Add test results for lexical transfer.)
m (Update "box" sentences given Zabby's fixes)
Line 9: Line 9:
 
== mto → eng evaluation (lexical transfer) ==
 
== mto → eng evaluation (lexical transfer) ==
  
* Monolingual coverage over mto.sentences.txt: 306/702 (~43.6%)
+
* Monolingual coverage over mto.sentences.txt: 310/702 (~44.2%)
* Bilingual coverage over mto.sentences.txt: 95/531 (~17.9%)
+
* Bilingual coverage over mto.sentences.txt: 99/531 (~18.7%)
  
 
=== Test of lexical transfer over 10 randomly selected sentences ===
 
=== Test of lexical transfer over 10 randomly selected sentences ===
  
 
<!-- line numbers from the corpus: 7, 9, 13, 14, 15, 16, 18, 20, 23, 26-->
 
<!-- line numbers from the corpus: 7, 9, 13, 14, 15, 16, 18, 20, 23, 26-->
 +
 
{|
 
{|
 
!width="25%"| mto
 
!width="25%"| mto
Line 27: Line 28:
 
|-
 
|-
 
| kajha juu’ müjit.
 
| kajha juu’ müjit.
| <code>^*kajha/*kajha$ ^*juu/*juu$' ^*müjit/*müjit$^.&lt;sent&gt;/.&lt;sent&gt;$^.&lt;sent&gt;/.&lt;sent&gt;$</code>
+
| <code>^kajha&lt;n&gt;&lt;sg&gt;/box&lt;n&gt;&lt;sg&gt;$ ^*juu/*juu$' ^*müjit/*müjit$^.&lt;sent&gt;/.&lt;sent&gt;$^.&lt;sent&gt;/.&lt;sent&gt;$</code>
| <code>*kajha *juu' *müjit.</code>
+
| <code>box *juu' *müjit.</code>
 
| Big boxes.
 
| Big boxes.
 
|-
 
|-
 
| juu’k tse’e ’anyükojmit jayu ntijumüt, tùùtük kyajha dü ’apiv’uts müjit.
 
| juu’k tse’e ’anyükojmit jayu ntijumüt, tùùtük kyajha dü ’apiv’uts müjit.
| <code>^*juu/*juu$'^*k/*k$ ^*tse/*tse$'^*e/*e$ '^*anyükojmit/*anyükojmit$ ^jäyu&lt;n&gt;&lt;sg&gt;&lt;px1&gt;/person&lt;n&gt;&lt;sg&gt;&lt;px1&gt;$ ^*ntijumüt/*ntijumüt$^,&lt;cm&gt;/,&lt;cm&gt;$ ^*tùùtük/*tùùtük$ ^*kyajha/*kyajha$ ^*dü/*dü$ '^*apiv/*apiv$'^*uts/*uts$ ^*müjit/*müjit$^.&lt;sent&gt;/.&lt;sent&gt;$^.&lt;sent&gt;/.&lt;sent&gt;$</code>
+
| <code>^*juu/*juu$'^*k/*k$ ^*tse/*tse$'^*e/*e$ '^*anyükojmit/*anyükojmit$ ^jäyu&lt;n&gt;&lt;sg&gt;/person&lt;n&gt;&lt;sg&gt;$ ^*ntijumüt/*ntijumüt$^,&lt;cm&gt;/,&lt;cm&gt;$ ^*tùùtük/*tùùtük$ ^*kyajha/*kyajha$ ^*dü/*dü$ '^*apiv/*apiv$'^*uts/*uts$ ^*müjit/*müjit$^.&lt;sent&gt;/.&lt;sent&gt;$^.&lt;sent&gt;/.&lt;sent&gt;$</code>
| <code>*juu'*k *tse'*e '*anyükojmit #person *ntijumüt, *tùùtük *kyajha *dü '*apiv'*uts *müjit.</code>
+
| <code>*juu'*k *tse'*e '*anyükojmit person *ntijumüt, *tùùtük *kyajha *dü '*apiv'*uts *müjit.</code>
 
| And the Totontepecanos, we could say for example, would fill six large boxes.
 
| And the Totontepecanos, we could say for example, would fill six large boxes.
 
|-
 
|-

Revision as of 22:31, 1 May 2022

Resources for machine translation between Mixe and English.

External resources

mto → eng evaluation (lexical transfer)

  • Monolingual coverage over mto.sentences.txt: 310/702 (~44.2%)
  • Bilingual coverage over mto.sentences.txt: 99/531 (~18.7%)

Test of lexical transfer over 10 randomly selected sentences

mto mto-eng-biltrans mto-eng eng
’ax paa’tyük tse?e jè vintsën vyaajny: '^*ax/*ax$ ^*paa/*paa$'^*tyük/*tyük$ ^*tse/*tse$^?<sent>/?<sent>$^*e/*e$ ^*jè/*jè$ ^vintsën<n><sg><px1>/owner<n><sg><px1>$ ^*vyaajny/*vyaajny$^:<sent>/:<sent>$^.<sent>/.<sent>$ '*ax *paa'*tyük *tse?*e *jè #owner *vyaajny: Now the boss said:
kajha juu’ müjit. ^kajha<n><sg>/box<n><sg>$ ^*juu/*juu$' ^*müjit/*müjit$^.<sent>/.<sent>$^.<sent>/.<sent>$ box *juu' *müjit. Big boxes.
juu’k tse’e ’anyükojmit jayu ntijumüt, tùùtük kyajha dü ’apiv’uts müjit. ^*juu/*juu$'^*k/*k$ ^*tse/*tse$'^*e/*e$ '^*anyükojmit/*anyükojmit$ ^jäyu<n><sg>/person<n><sg>$ ^*ntijumüt/*ntijumüt$^,<cm>/,<cm>$ ^*tùùtük/*tùùtük$ ^*kyajha/*kyajha$ ^*dü/*dü$ '^*apiv/*apiv$'^*uts/*uts$ ^*müjit/*müjit$^.<sent>/.<sent>$^.<sent>/.<sent>$ *juu'*k *tse'*e '*anyükojmit person *ntijumüt, *tùùtük *kyajha *dü '*apiv'*uts *müjit. And the Totontepecanos, we could say for example, would fill six large boxes.
jats viijnk jayu nay tùùjtük laata dü ’apiv’ùjtspa, pero mu… ’este pi’k muutskitse’e jüdü’üm jè laata. ^*jats/*jats$ ^*viijnk/*viijnk$ ^jäyu<n><sg><px1>/person<n><sg><px1>$ ^*nay/*nay$ ^*tùùjtük/*tùùjtük$ ^laata<n><sg><px1>/tin<n><sg><px1>$ ^*dü/*dü$ '^*apiv/*apiv$'^*ùjtspa/*ùjtspa$^,<cm>/,<cm>$ ^*pero/*pero$ ^*mu/*mu$^.<sent>/.<sent>$^.<sent>/.<sent>$^.<sent>/.<sent>$ '^*este/*este$ ^*pi/*pi$'^*k/*k$ ^*muutskitse/*muutskitse$'^*e/*e$ ^*jüdü/*jüdü$'^*üm/*üm$ ^*jè/*jè$ ^laata<n><sg><px1>/tin<n><sg><px1>$^.<sent>/.<sent>$^.<sent>/.<sent>$ *jats *viijnk #person *nay *tùùjtük #tin *dü '*apiv'*ùjtspa, *pero *mu... '*este *pi'*k *muutskitse'*e *jüdü'*üm *jè #tin. And the others would fill like six tins, but li-um, they were little, thus, those tins.
’ax ’akijpxak tse’e yakmujùùyda. '^*ax/*ax$ '^*akijpxak/*akijpxak$ ^*tse/*tse$'^*e/*e$ ^*yakmujùùyda/*yakmujùùyda$^.<sent>/.<sent>$^.<sent>/.<sent>$ '*ax '*akijpxak *tse'*e *yakmujùùyda. And they would pay them the same.
vanxüp ntijumüt, pën majk peesu ve’e yakmujùùyda y’ijt, tù’kkë’ëyji tse’e tum majk peesu. ^vanxüp<n><sg><px1>/amount<n><sg><px1>$ ^*ntijumüt/*ntijumüt$^,<cm>/,<cm>$ ^*pën/*pën$ ^*majk/*majk$ ^*peesu/*peesu$ ^*ve/*ve$'^*e/*e$ ^*yakmujùùyda/*yakmujùùyda$ ^*y/*y$'^*ijt/*ijt$^,<cm>/,<cm>$ ^*tù/*tù$'^*kkë/*kkë$'^*ëyji/*ëyji$ ^*tse/*tse$'^*e/*e$ ^*tum/*tum$ ^*majk/*majk$ ^*peesu/*peesu$^.<sent>/.<sent>$^.<sent>/.<sent>$ #amount *ntijumüt, *pën *majk *peesu *ve'*e *yakmujùùyda *y'*ijt, *tù'*kkë'*ëyji *tse'*e *tum *majk *peesu. We could say for example, if it was ten pesos they would get paid, everyone would be given ten pesos.
ka’amts jèm dü ‘akèèka-I jats nay vanxüp yakmujùùyvat ’ax jù’n jè’è juu’ pi’kit laata tü dü ’apiv’uts. ^*ka/*ka$'^*amts/*amts$ ^*jèm/*jèm$ ^*dü/*dü$ '^*akèèka/*akèèka$-^*I/*I$ ^*jats/*jats$ ^*nay/*nay$ ^vanxüp<n><sg><px1>/amount<n><sg><px1>$ ^*yakmujùùyvat/*yakmujùùyvat$ '^*ax/*ax$ ^*jù/*jù$'^*n/*n$ ^*jè/*jè$'^*è/*è$ ^*juu/*juu$' ^*pi/*pi$'^*kit/*kit$ ^laata<n><sg><px1>/tin<n><sg><px1>$ ^*tü/*tü$ ^*dü/*dü$ '^*apiv/*apiv$'^*uts/*uts$^.<sent>/.<sent>$^.<sent>/.<sent>$ *ka'*amts *jèm *dü '*akèèka-*I *jats *nay #amount *yakmujùùyvat '*ax *jù'*n *jè'*è *juu' *pi'*kit #tin *tü *dü '*apiv'*uts. It wasn’t fair that they would give them the same amount as the people who filled the little tins.
jè nde’m jayu dü nüüjma: ^*jè/*jè$ ^*nde/*nde$'^*m/*m$ ^jäyu<n><sg><px1>/person<n><sg><px1>$ ^*dü/*dü$ ^*nüüjma/*nüüjma$^:<sent>/:<sent>$^.<sent>/.<sent>$ *jè *nde'*m #person *dü *nüüjma: He said to the boss:
’ax, ’ax kum ’amaaxünts jèm jè jayu kyëts. '^*ax/*ax$^,<cm>/,<cm>$ '^*ax/*ax$ ^*kum/*kum$ '^*amaaxünts/*amaaxünts$ ^*jèm/*jèm$ ^*jè/*jè$ ^jäyu<n><sg><px1>/person<n><sg><px1>$ ^*kyëts/*kyëts$^.<sent>/.<sent>$^.<sent>/.<sent>$ '*ax, '*ax *kum '*amaaxünts *jèm *jè #person *kyëts. And the man, the man from here, he defends himself this way, struggling mightily anyway he could to express himself.
kuutse’e jèp ’ado’om mnëjkxta vintsotse’e jè jayu xjamëëda kotsjada? ^*kuutse/*kuutse$'^*e/*e$ ^*jèp/*jèp$ '^*ado/*ado$'^*om/*om$ ^*mnëjkxta/*mnëjkxta$ ^*vintsotse/*vintsotse$'^*e/*e$ ^*jè/*jè$ ^jäyu<n><sg><px1>/person<n><sg><px1>$ ^*xjamëëda/*xjamëëda$ ^*kotsjada/*kotsjada$^?<sent>/?<sent>$^.<sent>/.<sent>$ *kuutse'*e *jèp '*ado'*om *mnëjkxta *vintsotse'*e *jè #person *xjamëëda *kotsjada? When you travel there to the other side, how are you able to speak with those people?