Comparative Analysis of Tamil and English News Text Summarization Using Text Rank Algorithm

Abstract

The exponential growth of newsgroups has made it more difficult to gain accurate access to a large amount of data. To deal with the massive amounts of data, efficient and effective methods are needed. One such method is text summarization, which presents data in a condensed format. It would be beneficial for readers to be able to get a wide variety of news in a short amount of time if the news is simplified.In this article, we use the English Newsgroup datasets and the Tamil Newsgroup datasets to automate News Summaries using the text rank algorithm. The proposed work was created using a changed Text Rank algorithm based on the principle of word frequency. The suggested approach creates vectors of words as nodes and similarities between two words as the edge between them, which is the webbing between them. Term frequency assigns various weights to different terms in a sentence, while standard cosine similarity regards both of them similarly. The vector is rendered sparse and divided into clusters based on the premise that sentences inside a cluster are identical and sentences from various clusters reflect their dissimilarity. The performance assessment of the proposed summarization strategy in two types of Newsgroup datasets demonstrates its usefulness in terms of the accuracy parameter.