A Novel Sentence Ranking Clustering Based Algorithm for Text Summarization

Akanksha Sahu; Chander Diwaker

Abstract

Text Summarization is the process of generating reduced text from source document by extraction or generation. A novel approach will be presented for extractive text summarization. The proposed technique performs filtering and then ranking on sentences based on NLP text mining techniques. This method uses the synonyms extracted from online thesaurus dictionary to identify similar sentences and then clusters these sentences to summarize. The sentences are ranked based on sentence ranking algorithm and then the sentences from the filtered vectors which have a rank above a certain threshold value are included. A new method will be introduced where the user is allowed to manage the summary size as per his requirements. Unlike previous techniques, this technique takes into account synonyms of the keywords found in sentences and considers them while clustering. In the end, the created summaries by using this technique will be compared with others produced by commercial and research applications to demonstrate the efficiency of our technique.Cite this Article Akanksha Sahu, Chander Diwaker. A Novel Sentence Ranking Clustering Based Algorithm for Text Summarization. Journal of Artificial Intelligence Research & Advances. 2015; 2(2): 1–6p.

Abstract

Keywords