A Bayesian classifier for the identification of non-referential pronouns in Arabic

Machine learning has become the predominant problem-solving strategy for computational linguistics problems in the last decade. In this paper, we present an implemented machine learning system for the automatic identification of non-referential pronouns in Arabic texts. Our system is based on a Bayesian network which has shown its efficiency for modeling NLP problems. We have evaluated our approach on common data sets and we have obtained encouraging results, proving that the learning approach achieves accuracy better than the rule-based approach.