Survey on Graph and Cluster Based approaches in Multi-document Text Summarization

Abstract

In today's era of World Wide Web, on-line information is increasing exponentially day by day. So there is a need to condense corpus of documents into useful information automatically. Automatic Text summarization plays an important role to extract salient feature from corpus of documents, which helps user to get useful information in short time and less effort. Summarization reduces the complexity of a document while retaining its important features. Recently, most researchers have transferred their efforts from single to multi document summarization but they have to be aware of the issues of redundancy, sentence ordering, fluency, etc. There are wide varieties of approaches in Multi-document Text Summarization like Graph Based, Cluster Based, Time Based and Term frequency -Inverse document frequency Based etc. The survey starts introducing Multi-document text Summarization (MDS) and then discusses various methods of MDS which fall under the Graph and Cluster Based methods. In this paper, we have analysed Graph and Cluster Based methods proposed by various researchers in the field and we sort out some of the problems in applied procedures and also pin out advantages, which would help future researchers working in the area, to get significant instruction for further analysis. Using this information one can generate new or even hybrid methods in Multi-document summarization.