Lexical Disambiguation of Arabic Language: An Experimental Study

Abstract

"In this paper we test some supervised algorithmsthat most of the existing related works of word sensedisambiguation have cited. Due to the lack of linguistic data forthe Arabic language, we work on non-annotated corpus and withthe help of four annotators; we were able to annotate thedifferent samples containing the ambiguous words. Since that, wetest the Naïve Bayes algorithm, the decision lists and theexemplar based algorithm. During the experimental study, wetest the influence of the window size on the disambiguationquality, the derivation and the technique of smoothing for the(2n+1)-grams. For these tests the exemplar based algorithmachieves the best rate of precision."