Acquisition System for Arabic Noun Morphology

Abstract

Many papers have discussed different aspects of Arabic verb morphology. Some of them used patterns; others used patterns and affixes. But very few have discussed Arabic noun morphology particularly for nouns that are not derived from verbs. In this paper we describe a learning system that can analyze Arabic nouns to produce their morphological information and their paradigms with respect to both gender and number using a rule base that uses suffix analysis as well as pattern analysis. The system utilizes user-feedback to classify the noun and identify the group that it belongs to.