Text Summarization: Using Combinational Statistical And Linguistic Methods

Abstract

Summaries are considered as useful indicators of the document content. Traditionally, summaries are created by humans, through reading a text and identifying the most important points in it. But the information overloading makes such manual work very difficult. Many attempts have been made to automate the summarization task. This project aims at proposing a methodology for automatic text summarization based on extractive methods. Extractive summarization needs the selection of a subset of important, meaningful sentences from the source text. The work presents a method for automatic text summarization of a single computer science article using statistical and linguistic approach. About 50 experts' persons summaries were used to evaluate the proposed method. Our results show that using the combinational statistical and linguistic methods is better than using each method alone.