Comparative Summarization via Latent Dirichlet Allocation

Comparative Summarization via Latent Dirichlet Allocation

This paper aims to explore the possibility of using Latent Dirichlet Allocation (LDA) for multi-document comparative summarization which detects the main differences in documents. The first two sections of this paper focus on the definition of comparative summarization and a brief explanation of using the LDA topic model in this context. In the last three sections, our novel method for multi-document com- parative summarization using LDA is presented and also its results are compared with the results of a similar method based on Latent Semantic Analysis.

Keywords: comparative summarization, latent dirichlet allocation, latent semantic analysis, topic model

Year: 2013

Download: download Full text [356 kB]

Authors of this publication:

Michal Campr


Michal graduated from the University of West Bohemia in 2011, specialized in software engineering. He is interested in text summarization.

Karel Je┼żek

Phone:  +420 377632475

Karel is the former group coordinator and a supervisor of PhD students working at research projects of this Group.

Related Projects:


Automatic Text Summarisation

Authors:  Josef Steinberger, Karel Je┼żek, Michal Campr, Ji┼Ö├ş Hynek
Desc.:Automatic text summarisation using various text mining methods, mainly Latent Semantic Analysis (LSA).