Home // IMMM 2011, The First International Conference on Advances in Information Mining and Management // View article

Analyzing the Use of Word Graphs for Abstractive Text Summarization

Authors:
Elena Lloret
Manuel Palomar

Keywords: Human Language Technologies; automated retrieval and mining;automated content summarization; abstractive techniques; graph-based algorithms

Abstract:
This paper focuses on abstractive text summarization. Our aim is to explore to what extent new sentences generated employing a word graph-based method (which either compress or merge information) are suitable for producing abstracts. Moreover, in order to decide which of the new sentences should be included in the abstractive summary, an extractive text summarization approach is developed (i.e., COMPENDIUM, so that the most relevant abstractive sentences can be selected and extracted. As shown by the results obtained, this task is very challenging. However, preliminary experiments carried out prove that the combination of extractive and abstractive information is a more suitable strategy to adopt towards the generation of abstracts.

Pages: 61 to 66

Copyright: Copyright (c) IARIA, 2011

Publication date: October 23, 2011

Published in: conference

ISSN: 2326-9332

ISBN: 978-1-61208-162-5

Location: Barcelona, Spain

Dates: from October 23, 2011 to October 29, 2011