Title
Automatic alignment of news texts and their multi-document summaries: Comparison among methods
Date Issued
01 January 2014
Access level
open access
Resource Type
conference paper
Author(s)
University of São Paulo
Publisher(s)
Springer Verlag
Abstract
Aligning texts and their multi-document summaries is the task of determining the correspondences among textual segments in the texts and in their corresponding summaries. The study of alignments allows a better understanding of the multi-document summarization process, which may subsidize new summarization models for producing more informative summaries. In this paper, we investigate some approaches for text-summary sentence alignment, including superficial, deep and hybrid approaches. Our results show that superficial approaches may obtain very good results.
Start page
220
End page
231
Volume
8775
Language
English
OCDE Knowledge area
Bioinformática
Scopus EID
2-s2.0-84908564195
ISBN
9783319097602
Source
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Resource of which it is part
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN of the container
03029743
ISBN of the container
978-331909760-2
Conference
11th International Conference on Computational Processing of Portuguese, PROPOR 2014
Sources of information:
Directorio de Producción Científica
Scopus