The anaphors and ellipses resolution is a challenging task for most of NLP applications. However, few works have addressed this problem in Arabic. In fact the richness of the Arabic language and the scarcity of labeled data make the task more difficult. This paper offers a generic approach that deals with the identification and the resolution of both pronominal anaphors and ellipses in Arabic texts. To identify referential anaphors and ellipses positions, we propose a self-training SVM method based on a set of pattern-based and linguistic-based criteria. For the resolution step, we propose a novel hybrid method combining a reinforcement learning method with Word Embedding models. The reinforcement learning method uses an adapted version of Q-learning algorithm to finds the optimal combination of features. It exploits a set of morphological and syntactical features. The Word Embedding based method uses word representation models to check the semantic validity of the candidates. The evaluation of the identification approach gives a precision of 99.23% for pronominal anaphors and 94.33% for ellipses. The resolution approach reaches a precision of 80% for pronominal anaphors and 97.5% for ellipses. Our proposed approaches can be easily adapted to English and French to provide multilingual approaches.

This paper deals with the resolution of pronominal anaphora and zero anaphora in Arabic language. While researchers have treated the two phenomena separately, we propose a generic approach for both of them. Our resolution system combines a Q-learning reinforcement method and Word Embedding models. The Q-learning method uses syntactic criteria as preference factors to select candidate antecedents. It reinforces the best combination criteria for evaluating candidate antecedents. The Word Embedding models provide semantic similarity measures that help to validate the best antecedent. Our approach is evaluated on different type of Arabic texts and the obtained precision can reach79.37%.

The classification of pronouns as referential or non-referential is necessary for many NLP tasks. However, there are few works interested in this problem in the Arabic language. In this paper, we present a semi-supervised machine learning approach based on a Self-training SVM method for the identification of non-referential pronouns in the Arabic texts. A set of patterns-based and linguistic-based information is used as classification features in our machine learning system. The proposed Self-Training SVM algorithm includes three steps: training, prediction and selection step. It trains SVM classifier on a small set of labeled data, predicts labels of unlabeled data, selects the most accurate and the most informative newly labeled data and adds them to the training dataset. The selection step uses some geometric measures and analyses their relevance on the classification performance. The evaluation of our approach on the training and test data presents good results that can reach up to 96.85% of precision.

Anaphora pronoun resolution is an active task of Natural Language Processing (NLP) that is necessary in many NLP applications. It allows revealing the ambiguity, verifying the coherence and understanding the context. The aim of this paper is how to combine salience factors in order to rich a best pronoun resolution in Arabic texts. The factors allow to disambiguate candidate antecedents and finding out, for each pronoun, the most likely one from the list of all potential antecedents. For this, we explore two approaches that use the same set of salience factors. Our first approach is based on reinforcement learning using a Markov Decision Process (MDP). It involves the combination subsets of salience factors. The second is an adaptation of our first approach according to multilingual robust approach of Mitkov. The comparative study shows that the performance of the same set of factors can change according to the resolution process and context. The evaluation of the two approaches is conducted on a corpus of literary texts.