نحو بناء موارد لسانية لبلورة محلل تركيبي للجملة العربية

Abstract

The present paper is intended as a contribution to the automatic treatment of Arabic language. It presents a project conducted at the Institute for Studies and Research on Arabization to develop a Syntactic Analyzer for this language. Taking into account the characteristics of Arabic language syntax, our framework draws on both old and modern theories of Arabic linguistics to account for our twofold objective. The project aims at classifying the syntactic rules that govern the Arabic sentence typology and building a syntactic lexicon. 385 syntactic rules constituting Arabic language, including different syntactic patterns and sentence forms, have been elaborated. The syntactic lexicon is taken to be an open source that includes a database of 24171 entries tagged with combinatorial features (that contribute to specifying various verb categories and context of use) in a manual and semi-automatic way. Such classification allows to better analyze Arabic syntax and, thus, enhance the role of the lexicon as an important linguistic resource in the automated treatment of Arabic language in general, and its various applications in grammar analysis in particular.