Hybrid Proposed Method Using Statistical Linguistic Features (SLF) And Neural Network In Arabic Texts Summarization

Abstract

With the counting growth of electronic information on World Wide Web, it has become necessary important to provide mechanisms to find and present a shorter version would suffice, so that automatic text summarization technique which plays an important role to help users to determine whether it has to do with information they need or not. In this paper we present a short historical overview and advancement of automatic text summarization and the most relevant approaches currently used in this area. We proposed a new technique in automatic summarization area for Arabic news articles using a neural network by selecting important sentences from the original text and put it in the summary. A Multi-layer Perceptron neural network (MLP) is trained to learn the relevant characteristics of sentences that should be included in the summary of the article. The neural network is then modified to generalize and combine the relevant characteristics apparent in summary sentences . Finally, system is evaluated by comparing the final summary of the system with the summary produced by expert in Arabic language, we measure the performance of the system by computing the precision, recall and F-Measure and we obtain good result that we display in the conclusion.