Partitioned-based clustering approaches for single document extractive text summarization

Document Type

Conference Article

Publication Title

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Abstract

This article presents an extractive text summarization technique for single document using partition based clustering algorithms. Clustering of sentences is performed where the importance of each sentence in a document is attributed with three features namely, term score, keywords and average cosine similarity. Two clustering techniques, namely, k-means and fuzzy C-means are considered. To generate the summary, sentences are selected using two similarity calculation methods and the results are obtained for different compression rates (20%–60%). The results are quite promising with respect to the references used for evaluation.

First Page

297

Last Page

307

DOI

10.1007/978-3-319-71928-3_29

Publication Date

1-1-2017

This document is currently not available here.

Share

COinS