TextSum: An Automatic Text Summarization Tool using Bayesian Classification

Abstract

Automatic text summarization is the process of summarizing text from one or more documents in order to help the reader to determine if [s]he has to read the document(s) in full. Although automatic summarization is around for many years, there is a renewed interest in this area from the government, industry, and academia. The reason is the availability of large quantities of information due to world wide web and the necessity to summarize such information. Automatic summarization involves reducing text document(s) into shorter form which conveys the same meaning as the original documents. Summarization may be done by extractive methods or abstractive methods. This paper consists of an Automatic Text Summarization Tool which takes into account several features, including sentence position, positive and negative key words, sentence centrality etc., and employs the Bayesian classification algorithm to identify the sentences that are to be included in the summary.