A Rule Based Persons Names Arabic Extraction System

Communications of the IBIMA

Download PDF

Ali Elsebai1, Farid Meziane2 and Fatma Zohra Belkredim3

 

1,2School of Computing, Science and Engineering, University of Salford Salford M5 4WT, UK

3Departement d’Informatique, Universitie Hassiba Ben Bouali Chlef, Algeria

Volume 2009 (2009), Article ID 730570, Communications of the IBIMA, 7 pages, DOI:

Received date : 24 July 2009; Accepted date : 23 October 2009; Published date : 15 December 2009

Abstract

Named Entity Extraction is a very new in Arabic Natural Language processing although it has reached maturity for some other languages such as English and French. In this paper, we describe the development and implementation of a person name named entity recognition system for the Arabic Language. We adopt a rule based approach make used of the output produced by the Buckwalter Arabic Morphological Analyser (BAMA),  and we used a set of keywords to guide us to the phrases that probably include person names. We have also compared our system with (PERA) Person Name Entity Recognition for Arabic [9] which is based on a lexicon, in the form of gazetteer name lists, and a grammar, in the form of regular expressions. Our system achieves an F-measure of 89% which is an improvement on the results reported by (PERA).

Keywords: Message Understanding Conference (MUC), Arabic Named Entity.