Difference between revisions of "Japhug and English/Structural Transfer"

From LING073
Jump to: navigation, search
(→‎Pre-Evaluation)
(→‎Pre-Evaluation)
Line 21: Line 21:
 
</pre>
 
</pre>
  
−
'''Coverage Test in Bilingual Transducer'''
+
'''Coverage Test on jya.tests.txt in Bilingual Transducer'''
 
<pre>
 
<pre>
 
TOP UNKNOWN WORDS:
 
TOP UNKNOWN WORDS:
−
    60 ^ndɤre/*ndɤre$
+
      1 ^ʁnɯtɤrtsɯɣ/*ʁnɯtɤrtsɯɣ$
−
    43 ^ma/*ma$
+
      1 ^ɲɯrganɯ/*ɲɯrganɯ$
−
    28 ^ti/*ti$
+
      1 ^ɲɤmtsɯr/*ɲɤmtsɯr$
−
    21 ^toti/*toti$
+
      1 ^ɯʑo/*ɯʑo$
−
    19 ^tɕheme/*tɕheme$
+
      1 ^ɯtaʁ/*ɯtaʁ$
−
    17 ^mbro/*mbro$
+
      1 ^ɯjme/*ɯjme$
−
    16 ^nɯnɯ/*nɯnɯ$
+
      1 ^ɕɤrɯ/*ɕɤrɯ$
−
    13 ^qachɣa/*qachɣa$
+
      1 ^xuexiao/*xuexiao$
−
    12 ^tʂu/*tʂu$
+
      1 ^tʂu/*tʂu$
−
    12 ^ni/*ni$
+
      1 ^pɯ́wɣsɯmtoa/*pɯ́wɣsɯmtoa$
−
    12 ^aʑo/*aʑo$
+
      1 ^mɯpɯrɤʑia/*mɯpɯrɤʑia$
−
    11 ^iɕqha/*iɕqha$
+
       1 ^mɯpjɤpenɯ/*mɯpjɤpenɯ$
−
    10 ^li/*li$
+
       1 ^mɤpe/*mɤpe$
−
       9 ^ɯʑo/*ɯʑo$
+
       1 ^kɯkɯra/*kɯkɯra$
−
       9 ^kha/*kha$
+
       1 ^kɤzo/*kɤzo$
−
       8 ^rɟɤlpu/*rɟɤlpu$
+
       1 ^khramba/*khramba$
−
       8 ^rcanɯ/*rcanɯ$
+
       1 ^aʑo/*aʑo$
−
       8 ^kɯwxti/*kɯwxti$
+
       1 ^ambro/*ambro$
−
       7 ^ɯŋgɯ/*ɯŋgɯ$
+
coverage: 24 / 42 (~0.57142857142857142857)
−
       7 ^ɯrkɯ/*ɯrkɯ$
+
remaining unknown forms: 18
−
coverage: 1862 / 3143 (~0.59242761692650334076)
+
ambiguity: 54 / 42 (~1.28571428571428571429)  
−
remaining unknown forms: 1281
 
−
ambiguity: 3632 / 3143 (~1.15558383709831371301)
 
 
</pre>
 
</pre>
  
−
'''Coverage Test in Monolingual Transducer'''
+
'''Coverage Test on jya.tests.txt in Monolingual Transducer'''
 
<pre>
 
<pre>
 
TOP UNKNOWN WORDS:
 
TOP UNKNOWN WORDS:
−
       8 ^rɟɤlpu/*rɟɤlpu$
+
       1 ^ʁnɯtɤrtsɯɣ/*ʁnɯtɤrtsɯɣ$
−
       8 ^rcanɯ/*rcanɯ$
+
       1 ^ɲɯrganɯ/*ɲɯrganɯ$
−
       7 ^ɯrkɯ/*ɯrkɯ$
+
       1 ^ɯjme/*ɯjme$
−
       7 ^tɤwa/*tɤwa$
+
       1 ^xuexiao/*xuexiao$
−
       7 ^tɤsno/*tɤsno$
+
       1 ^pɯ́wɣsɯmtoa/*pɯ́wɣsɯmtoa$
−
       7 ^tɕɤn/*tɕɤn$
+
       1 ^mɯpɯrɤʑia/*mɯpɯrɤʑia$
−
       7 ^ki/*ki$
+
       1 ^mɯpjɤpenɯ/*mɯpjɤpenɯ$
−
       6 ^ɯpi/*ɯpi$
+
       1 ^kɯkɯra/*kɯkɯra$
−
      6 ^ɣɯ/*ɣɯ$
+
coverage: 33 / 41 (~0.80487804878048780488)
−
      6 ^tɤtɕɯ/*tɤtɕɯ$
+
remaining unknown forms: 8
−
      6 ^qhe/*qhe$
+
ambiguity: 82 / 41 (2)  
−
      5 ^tɯɣjɤn/*tɯɣjɤn$
 
−
      5 ^ndɤ/*ndɤ$
 
−
      5 ^kɯsthɯci/*kɯsthɯci$
 
−
      5 ^cho/*cho$
 
−
      4 ^tuβzea/*tuβzea$
 
−
      4 ^tinɯ/*tinɯ$
 
−
      4 ^nɯnaj/*nɯnaj$
 
−
      4 ^kɯkɯtu/*kɯkɯtu$
 
−
      3 ^ɲɯsɯsɤm/*ɲɯsɯsɤm$
 
−
coverage: 2863 / 3339 (~0.85744234800838574423)
 
−
remaining unknown forms: 476
 
−
ambiguity: 6101 / 3339 (~1.82719377058999700509)
 
 
</pre>
 
</pre>
  

Revision as of 00:31, 3 May 2023


Pre-Evaluation

Statistics About Input Files

Number of words in reference: 47
Number of words in test: 33
Number of unknown words (marked with a star) in test: 18
Percentage of unknown words: 54.55 %

Results when removing unknown-word marks (stars)

Edit distance: 47
Word error rate (WER): 100.00 %
Number of position-independent correct words: 0
Position-independent word error rate (PER): 100.00 %

Coverage Test on jya.tests.txt in Bilingual Transducer

TOP UNKNOWN WORDS:
      1 ^ʁnɯtɤrtsɯɣ/*ʁnɯtɤrtsɯɣ$
      1 ^ɲɯrganɯ/*ɲɯrganɯ$
      1 ^ɲɤmtsɯr/*ɲɤmtsɯr$
      1 ^ɯʑo/*ɯʑo$
      1 ^ɯtaʁ/*ɯtaʁ$
      1 ^ɯjme/*ɯjme$
      1 ^ɕɤrɯ/*ɕɤrɯ$
      1 ^xuexiao/*xuexiao$
      1 ^tʂu/*tʂu$
      1 ^pɯ́wɣsɯmtoa/*pɯ́wɣsɯmtoa$
      1 ^mɯpɯrɤʑia/*mɯpɯrɤʑia$
      1 ^mɯpjɤpenɯ/*mɯpjɤpenɯ$
      1 ^mɤpe/*mɤpe$
      1 ^kɯkɯra/*kɯkɯra$
      1 ^kɤzo/*kɤzo$
      1 ^khramba/*khramba$
      1 ^aʑo/*aʑo$
      1 ^ambro/*ambro$
coverage: 24 / 42 (~0.57142857142857142857)
remaining unknown forms: 18
ambiguity: 54 / 42 (~1.28571428571428571429) 

Coverage Test on jya.tests.txt in Monolingual Transducer

TOP UNKNOWN WORDS:
      1 ^ʁnɯtɤrtsɯɣ/*ʁnɯtɤrtsɯɣ$
      1 ^ɲɯrganɯ/*ɲɯrganɯ$
      1 ^ɯjme/*ɯjme$
      1 ^xuexiao/*xuexiao$
      1 ^pɯ́wɣsɯmtoa/*pɯ́wɣsɯmtoa$
      1 ^mɯpɯrɤʑia/*mɯpɯrɤʑia$
      1 ^mɯpjɤpenɯ/*mɯpjɤpenɯ$
      1 ^kɯkɯra/*kɯkɯra$
coverage: 33 / 41 (~0.80487804878048780488)
remaining unknown forms: 8
ambiguity: 82 / 41 (2) 

Post-Evaluation

Statistics About Input Files

 

Results when removing unknown-word marks (stars)