An Intelligent Framework for Natural Language Stems Processing

Abstract

This work describes an intelligent framework that enables the derivation of stems from inflected words Word stemming is one of the most important factors affecting the performance of many language applications including parsing syntactic analysis speech recognition retrieval systems medical systems tutoring systems biological systems and translation systems Computational stemming is essential for dealing with some natural language processing such as Arabic Language since Arabic is a highly inflected language Computational stemming is an urgent necessity for dealing with Arabic natural language processing The framework is based on logic programming that creates a program to enabling the computer to reason logically This framework provides information on semantics of words and resolves ambiguity It determines the position of each addition or bound morpheme and identifies whether the inflected word is a subject object or something else Position identification expression is vital for enhancing understandability mechanisms The proposed framework adapts bi-directional approaches It can deduce morphemes from inflected words or it can build inflected words from stems The proposed framework handles multi-word expressions and identification of names The framework is based on definiteclause grammar where rules are built according to Arabic patterns templates using programming language prolog as predicates in first-order logic This framework is based on using predicates in firstorder logic with object-oriented programming convention which can address problems of complexity This complexity of natural language processing comes from the huge amount of storage required This storage reduces the efficiency of the software system In order to deal with this complexity the research uses Prolog as it is based on efficient and simple proof routines It has dynamic memory allocation of automatic garbage collection This facility in addition to relieve the