Abstract
Position information has been proved to be very effective in document summarization, especially in generic summarization. Existing approaches mostly consider the information of sentence positions in a document, based on a sentence position hypothesis that the importance of a sentence decreases with its distance from the beginning of the document. In this paper, we consider another kind of position information, i.e., the word position information, which is based on the ordinal positions of word appearances instead of sentence positions. An extractive summarization model is proposed to provide an evaluation framework for the position information. The resulting systems are evaluated on various data sets to demonstrate the effectiveness of the position information in different summarization tasks. Experimental results show that word position information is more effective and adaptive than sentence position information.
Original language | English |
---|---|
Title of host publication | Coling 2010 - 23rd International Conference on Computational Linguistics, Proceedings of the Conference |
Pages | 919-927 |
Number of pages | 9 |
Publication status | Published - 1 Dec 2010 |
Event | 23rd International Conference on Computational Linguistics, Coling 2010 - Beijing, China Duration: 23 Aug 2010 → 27 Aug 2010 |
Conference
Conference | 23rd International Conference on Computational Linguistics, Coling 2010 |
---|---|
Country/Territory | China |
City | Beijing |
Period | 23/08/10 → 27/08/10 |
ASJC Scopus subject areas
- Language and Linguistics
- Computational Theory and Mathematics
- Linguistics and Language