تمثیل قواعد اللغة العربیة كحدود منطقیة في أنظمة معالجة اللغات الطبیعیة

Abstract

Usually, the natural language grammar is represented as a rules inside theprogram of natural language processing (NLP), but with a language such as theArabic language that have a flexible syntax, using this representation will producea large number of these rules that lead to increase the search time required and toincrease the size of the program, so in this paper we proposed a new representationmethod. We represent the Arabic language grammar as logical terms in theknowledge base, and we use the B+tree to store these terms since it is suitable withlarge data, so we reach the required syntactical term in a little time by using thesuitable searching method that b+tree provide it.In this paper we choose the Augmented Transition Network(ATN) to representthe language grammar since it is the most suitable parsing technique for a languagesuch as the Arabic language.