LexRank Algorithm: Application in Emails and Comparative Analysis

Aviva Munshi, Anoushka Mehra, Ashna Choudhury

email summarization, LexRank Algorithm, natural language processing, rogue scores

Text summarization can be described as the process that helps to shorten long pieces of text, with the goal of producing succinct and factual content that places specific focus on the basics present in the document. This is a known issue in machine learning and natural language processing, and the amount of attention given to it has only increased over many years, keeping in mind that there are copious quantities of data online. It also has the ability to collect useful information that can be managed fairly easily by humans and could be used for a wide range of purposes, such as text assessment. In this paper, we are attempting to present an automated text summary method that relies on LexRank Algorithm to find the most significant and appropriate statements in the long input text and make them a part of the short summary. In this project, given a set of data for a particular topic, the appropriate summary is produced using the LexRank algorithm. It is also capable of summarizing a single data as well. We are using college circulars as the data for testing the relevance of the produced summaries. We are also testing its relevance by testing the already available data by generating the ROUGE scores where the automatically generated summaries are compared with the manually written summaries of the same. 

https://doi.org/10.31871/IJNTR.7.5.9