Design of Syntactical Morphological and Semantical Analyzer (SMSA) for Processing Arabic Texts

摘要

This research describes an ongoing expert system developed for Natural Language Processing (NLP), and presents an approach for Arabic language manipulation, which integrates Syntactical Morphological and Semantical Analyzer (SMSA). Informational goal of SMSA is processing Arabic language from its inductive database, which is organized without dictionary. Search engine is built up as a high performance linguistic engine that facilitates analyses of written texts in Arabic language, performs full linguistic processing on text, and generates robust parser for Arabic sentences. Learned knowledge is represented in form of rules and facts. Reasoning and inference are accredited to aid grammar induction containing syntactical, morphological and semantical rules, which are conducive towards language processing. Arabic sentences are divided into two main parts: statement sentence and composition sentence. In turn, statement sentence is divided into a noun, verbal, and conditional sentences. Composition sentence is splited up into imperative and expletive composition. Words are divided into two parts: possession and functional words, also verbs and nouns are sorted in semantical groups. Programming language Visual Prolog 6.1 is used. Database is built up to save Arabic sentences frames as facts.