Difference between revisions of "Chechen/Disambiguation"

Revision as of 04:14, 8 March 2019

Initial Evaluation of Ambiguity

This is an evaluation of the basic corpus.

• Number of tokens: 2267
• Number of analyses: 2281
• Level of ambiguity: 0.99386

Ambiguity Identified

• зенhas two different readings:

• зен<n><cl_d><abs><sg> "loss, harm"
• зен<v><tv><inf> "examine"

• Two sentences with two different readings:

• ^доккху/доккху<v><tv><pres>$ ^зен/зен<v><tv><inf>/зен<n><cl_d><abs><sg>$ ^дан/дан<v><tv><inf>$ ^йиш/йиш<n><cl_j><abs><sg>$ ^ю/ю<v><cpl><cl_j>$^../..<sent>$
• ^Ас/ас<prn><pers><p1><erg><sg>$ ^аьлира/ала<v><iv><past><rem><wit>$ ^цуьнга/цуьнга<prn><pers><p3><all><sg>$ ^мотор/мотор<n><cl_u><abs><sg>$ ^зен/зен<v><tv><inf>/зен<n><cl_d><abs><sg>$^../..<sent>$

• This sentence is made out of an example sentence that contains an infinite verb in the grammar book and a phrase with another infinite verb from the dictionary.

Final Evaluation of Ambiguity

• Number of tokens: 2267
• Number of analyses: 2281
• Level of ambiguity: 0.99386

• Since the ambiguity is not identified from the corpus but from the dictionary, the level of ambiguity does not go up.