Uyghur/Transducer

Github Repository: [1]

Analyzer Evaluation

Total number of stems in transducer:

Counts for individual lexicons: 25

• Noun-Stems: 9
• Plural: 1
• Cases: 10
• Personal-Pronoun: 9
• Demonstrative-Pronoun: 7

Current coverage:

• coverage: 48 / 848 (~0.05660377358490566038)

Current ambiguity:

• 904 / 848 (~1.06603773584905660377)

List of unknown words:

TOP UNKNOWN WORDS:

    27 ^ۋە/*ۋە$
    24 ^خۇدا/*خۇدا$
    14 ^بىلەن/*بىلەن$
    13 ^خىتاي/*خىتاي$
    13 ^بولدى/*بولدى$
    11 ^ھەر/*ھەر$
    11 ^خىل/*خىل$
    10 ^يەنە/*يەنە$
     9 ^دېدى/*دېدى$
     8 ^خۇدانىڭ/*خۇدانىڭ$
     7 ^يەنى/*يەنى$
     7 ^ياخشى/*ياخشى$
     7 ^كېچە/*كېچە$
     7 ^كۈندۈز/*كۈندۈز$
     7 ^كۆردى/*كۆردى$
     7 ^قىلىپ/*قىلىپ$
     7 ^ئىكەنلىكىنى/*ئىكەنلىكىنى$
     6 ^شۇنداق/*شۇنداق$
     6 ^سوت/*سوت$
     6 ^جەمئىيىتى/*جەمئىيىتى$

Tests that Pass:

Corpus 1 of 23: commonwords

 4/4 (100.0%) tests pass (1/4 (25.0%) match gold)

Corpus 4 of 23: Personal Pronouns-morph

 9/9 (100.0%) tests pass (8/9 (88.89%) match gold)

Corpus 5 of 10: Plurals-morph

 4/4 (100.0%) tests pass (4/4 (100.0%) match gold)

Corpus 6 of 23: Demonstrative Pronouns-morph

 7/7 (100.0%) tests pass (7/7 (100.0%) match gold)

Corpus 8 of 23: Nouns-morph

 40/40 (100.0%) tests pass (40/40 (100.0%) match gold)

Generator Evaluation

Initial evaluation of morphological generation

Analyzer Test

Corpus 1 of 10: Personal Pronouns-morph

 9/9 (100.0%) tests pass (8/9 (88.89%) match gold)

Corpus 2 of 10: Demonstrative Pronouns-morph

 7/7 (100.0%) tests pass (7/7 (100.0%) match gold)

Corpus 3 of 10: Nouns-morph

 40/40 (100.0%) tests pass (40/40 (100.0%) match gold)

Corpus 5 of 10: Plurals-morph

 4/4 (100.0%) tests pass (4/4 (100.0%) match gold)

All tests pass.

Coverage

coverage: 48 / 848 (~0.05660377358490566038) remaining unknown forms: 800 ambiguity: 904 / 848 (~1.06603773584905660377)

Gen Test

Corpus 1 of 10: Personal Pronouns-gen

 9/9 (100.0%) tests pass (5/9 (55.56%) match gold)

Corpus 2 of 10: Demonstrative Pronouns-gen

 7/7 (100.0%) tests pass (7/7 (100.0%) match gold)

Corpus 3 of 10: Nouns-gen

 40/40 (100.0%) tests pass (12/40 (30.0%) match gold)

Corpus 5 of 10: Plurals-gen

 4/4 (100.0%) tests pass (0/4 (0.0%) match gold)

Final Assessment

Using ten twol rules:

Gen Test

Corpus 1 of 10: Personal Pronouns-gen

 9/9 (100.0%) tests pass (5/9 (55.56%) match gold)

Corpus 2 of 10: Demonstrative Pronouns-gen

 7/7 (100.0%) tests pass (7/7 (100.0%) match gold)

Corpus 3 of 10: Nouns-gen

 40/40 (100.0%) tests pass (27/40 (67.5%) match gold)

Corpus 4 of 10: Possessive Suffixes-gen

 2/2 (100.0%) tests pass (0/2 (0.0%) match gold)

Corpus 5 of 10: Plurals-gen

 4/4 (100.0%) tests pass (4/4 (100.0%) match gold)

Total: 71.67% Gold

Coverage

 coverage: 51 / 848 (~0.06014150943396226415)
 remaining unknown forms: 797
 ambiguity: 899 / 848 (~1.06014150943396226415)