Multi-document summarization

About: Multi-document summarization is a research topic. Over the lifetime, 2270 publications have been published within this topic receiving 71850 citations.

An Algorithm for Automatic Content Summarization in Modern Greek Language

Abstract: In this paper, basic principles are presented and an algorithm for extracting document content summary for Greek language, by using statistic method. The algorithm presented is based on grammatical rules and semantic information dedicated for Greek language. The algorithm has been tested on a variety of news articles and produces satisfactory results for a variety of thematic subjects. Apart from the algorithm, a test case is presented to validate its performance

A Learning Technique to Determine Criteria for Multiple Document Summarization

Abstract: In this paper we describe a new method of automatic summarization based on a learning step to identify criteria that maximize the correlation between human summary and peer extract. The proposed method uses a genetic algorithm to produce extracts from a collection of source documents describing the same event. Theses extracts are compared to human summaries using “Rouge measure” in order to identify the correlation between statistical and linguistic criteria and “Rouge score”. The experiment Results are presented for a document set extracted from the DUC’06 evaluation conference.

Automatic summarization for text based on mutual information

Abstract: Nowadays most of automatic summarization technologies often process a text as a whole with vector space model.These methods ignore the correlation degree among paragraphs,which lead to the inaccuracy of the summary that does not match with the main point of the text topics.In order to solve this problem,the mutual information theory is introduced to automatic summarization.The correlation degree among words,sentences and paragraphs is computed with an improved mutual information formula.Then the whole text is divided into smaller units which belong to different topics,and the topic sentences are extracted from the units by using an improved sentence weight measurement method.Finally,the summary is generated from the different topic sentences.The experimental results prove the validity of the proposed method in achieving the satisfactory results on automatic summarization.

Croatian text summarizer (CROSUM)

Abstract: The paper describes automatic summarization of the scientific papers in Croatian language. The goal of the CROSUM is to generate extracts with high percent of extract- worthiness and about the same size as the author's abstract. This preliminary research shows that extracts generated using the lemmatized wordforms dictionary are not quite different from extracts that are given on the base of the non-lemmatized wordforms dictionary. The research brought us to conclusion that we should develop a technique for identifying cue phrases from training corpus or some linguistic technique in order to improve the text summarization for Croatian language.

Multilingual text summarization with UNL

Abstract: The paper presents an algorithm for single document text summarization. Text Summarization is extraction of important information from the large documents by neglecting the unnecessary information with the help of different algorithms and providing the summary in compressed way which is very useful. Since UNL is language independent i.e. summarization with UNL is carried out by analyzing and removing the unnecessary relations. UNL is used in text summarization due to the reason that user can get the summary in his own native language.