Central Kurdish/Disambiguation

From LING073
Revision as of 19:16, 8 April 2021 by Rkamal1 (talk | contribs) (Constraint Example with Sentences)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search



Initial Evaluation of Ambiguity

  • Ambiguity in the transducer: ~1.07
  • Initial ambiguity in the corpus (Wikipedia): ~1.045

Constraint Example with Sentences

One source of ambiguity came from the ی suffix at the end of nouns. The suffix could be analyzed either as an izafa or as a personal suffix for a 3rd-person singular subject. The two sentences below demonstrate the successful disambiguation between the two forms in two specific scenarios. Both examples are contrived (not from an external corpus), though since the programmer is a native speaker of Sorani Kurdish it can essentially be considered authentic.

  • "سەگی گەورە" (big dog)

‌Here one would expect the ی to output as an izafa, because it is followed by a noun. The output was verified to be: ^سەگ<n><def><sg>+ی<pr>$

  • "ئەو سەگەکەی هێنا" (he brought the dog)

Here the ێ should be analyzed as a personal suffix, because it is immediately followed by a transitive verb. The output was: ^سەگ<n><def><sg>+ئەو<prn><pers><p3><sg>$

Final Evaluation of Ambiguity

  • Final ambiguity in the corpus (Wikipedia): ~1.029 (-35%)