Automated Bengali Document Summarization by Collaborating Individual Word & Sentence Scoring

Abstract

Bengali documents are increasing on the World Wide Web and it is becoming a overwhelming problem for the increasing large number of web users to reviewing and reduce the information. Many researches have been conducted in the field of Natural Language Processing for English documents and in order to serve with satisfactory accuracy. This research work proposed a simple and powerful extraction based method for summarizing of the Bengali text documents. The system could summarize a single document at a time. The ultimate objective of the proposed methodology helps readers to get summary and insight of the Bengali documents without reading revealing the in-depth details. In the proposed Bengali documents summary generation method there are four features: Preprocessing, Sentence Ranking and Summarization, Combining Parameters for Sentence Ranking, Summary Generator. The results of performance evaluation show that the average scores of Precision, Recall and final scores are 0.80, 0.67, and 0.72 respectively.