Difference between revisions of "Kperalt1/Final Project"

From LING073
Jump to: navigation, search
(→‎Shanghainese wuu → English eng Sentence Evaluation)
(→‎Coverage)
Line 62: Line 62:
 
** full translation output: #one 、 #two 、 #three 、 #four 、 #five 、 #six 、 #seven 、 #eight 、 #nine 、 #ten .
 
** full translation output: #one 、 #two 、 #three 、 #four 、 #five 、 #six 、 #seven 、 #eight 、 #nine 、 #ten .
  
−
==Coverage==
+
==Statistics about input files==
−
*TOP UNKNOWN WORDS:
+
-------------------------------------------------------
−
*coverage: 1 / 1 (1)
+
Number of words in reference: 8339
−
*remaining unknown forms: 0
+
Number of words in test: 10077
−
*ambiguity: 1 / 1 (1)
+
Number of unknown words (marked with a star) in test: 5756
 +
Percentage of unknown words: 57.12 %
 +
 
 +
Results when removing unknown-word marks (stars)
 +
-------------------------------------------------------
 +
Edit distance: 10630
 +
Word error rate (WER): 127.47 %
 +
Number of position-independent correct words: 49
 +
Position-independent word error rate (PER): 120.25 %
 +
 
  
 
[[Category:Shanghainese]]
 
[[Category:Shanghainese]]

Revision as of 15:46, 5 May 2025

Resources for machine translation between Shanghainese and English

  • Resources developed for Shanghainese:
    • Link to resources developed for Shanghainese: [1]
  • Resources developed for English:

External Resources

  • Link to repository language pair:[2]
  • Link to Shanghainese repository transducer:[3]
  • Link to English repository transducer:[4]
  • Link to Shanghainese to English Text: [5]

Shanghainese wuu → English eng Sentence Evaluation

Sentences

  • (1) wuu: 我爱侬。
    • eng: I love you.
    • lexical transfer output: ^我<prn><pers><p1><sg>/I<prn><pers><p1><sg>$ ^爱<vblex>/love<vblex>$ ^侬<prn><pers><p2><sg>/you<prn><pers><p2><sg>$ ^.<sent>/.<sent>
    • full translation output: #I #love #you .
  • (2) wuu: 我有很多朋友。
    • eng: I have many friends.
    • lexical transfer output: ^我<prn><pers><p1><sg>/I<prn><pers><p1><sg>$ ^有<vblex>/have<vblex>$ ^很多<adj>/many<adj>$ ^朋友<n>/friends<n><pl>$ ^.<sent>/.<sent>
    • full translation output: #I #have many #friends .
  • (3) wuu: 侬欢喜唱歌伐?
    • eng: Do you like singing?
    • lexical transfer output:
    • full translation output:
  • (4) wuu: 大家早上好。
    • eng: Good morning, everybody.
    • lexical transfer output:
    • full translation output:
  • (5) wuu: 侬要吃茶还是吃咖啡啊?
    • eng: Do you want to drink tea or coffee?
    • lexical transfer output:
    • full translation output:
  • (6) wuu: 是我个,勿是伊个。
    • eng: It's mine, not his.
    • lexical transfer output:
    • full translation output:
  • (7) wuu: 搿是我个行李。
    • eng: Here is my baggage.
    • lexical transfer output: ^搿<adj>/here<adj>$ ^是<vaux>/is<vaux>$ ^我<prn><pers><p1><sg><pos>/my<prn><pers><p1><sg><pos>/mine<prn><pers><p1><sg><pos>$ ^行李<n>/baggage<n>$ ^.<sent>/.<sent>
    • full translation output: #here #is #my #baggage .
  • (8) wuu: 伊是法国人。
    • eng: She is French. or He is French.
    • lexical transfer output: ^伊<prn><pers><p3><sg>/he<prn><pers><p3><sg><m>/it<prn><pers><p3><sg><nt>/her<prn><pers><p3><sg><f>/him<prn><pers><p3><sg><m>/she<prn><pers><p3><sg><f>$ ^是<vaux>/is<vaux>$ ^法国人<n>/French<n>$ ^.<sent>/.<sent>
    • full translation output: #he #is #French .
  • (9) wuu: 我每天六点钟起床。
    • eng: I get up at six every day.
    • lexical transfer output: ^我<prn><pers><p1><sg>/I<prn><pers><p1><sg>$ ^每天<adv>/every day<adv>$ ^六点钟<n>/six<n>$ ^起床<vblex>/wake up<vblex>$ ^.<sent>/.<sent>
    • full translation output: #I every day #six #wake up .
  • (10) wuu: 一、二、三、四、五、六、七、八、九、十。
    • eng: One, two, three, four, five, six, seven, right, nine, ten.
    • lexical transfer output: ^一<num>/one<num>$ 、 ^二<num>/two<num>$ 、 ^三<num>/three<num>$ 、 ^四<num>/four<num>$ 、 ^五<num>/five<num>$ 、 ^六<num>/six<num>$ 、 ^七<num>/seven<num>$ 、 ^八<num>/eight<num>$ 、 ^九<num>/nine<num>$ 、 ^十<num>/ten<num>$ ^.<sent>/.<sent>
    • full translation output: #one 、 #two 、 #three 、 #four 、 #five 、 #six 、 #seven 、 #eight 、 #nine 、 #ten .

Statistics about input files


Number of words in reference: 8339 Number of words in test: 10077 Number of unknown words (marked with a star) in test: 5756 Percentage of unknown words: 57.12 %

Results when removing unknown-word marks (stars)


Edit distance: 10630 Word error rate (WER): 127.47 % Number of position-independent correct words: 49 Position-independent word error rate (PER): 120.25 %