Automatic Summarization System For Arabic Texts

Muhajer Kareem, Suhad; Malik Dumamad, Yusra

	Automatic Summarization System For Arabic Texts
basrah journal of science
Article 1, Volume 30, Issue 1, 2012, Pages 63-75 PDF (0 K)
Authors
Suhad Muhajer Kareem; Yusra Malik Dumamad
Abstract
Automatic summarization plays an important role in light of the development of World Wide Web, which represents the most important ways to manage the large amount of text information that a user needs to read the text and by reducing the amount of text to help determine whether it has to do with the information they need or not. In this paper we proposed the automatic system for the Arabic text, trying to imitate human behavior so in writing the summary of the text provided. In this paper deals with method of a hybrid (statistical, linguistic), where we proposed several mathematical equations that merging a set of text properties to process the original text and get a find a summary coherent that contain useful information then was selected its efficient in the production of a summary is good and using the algorithm MMR has been removed redundant from sentences Summary produced. And then evaluated by comparing the fin Summary of the system with the summaries produced by experts in the Arabic language using ROUGE evaluation system by computing the precision, recall and F-Measure. The results were good and very encouraging , note that the systems summary of the Arabic language are very small compared with the rest of the languages. The system is programmed in Delphi 7 language.
Keywords
MMR; measure

Statistics Article View: 73 PDF Download: 34