Abstract
Multi-document summarization aims to extract information from the original multiple unstructured text documents or other types of multimedia about the same topic and this paper will focus on eventbased multi-document summarization. Event-based extractive summarization attempts to extract sentences and re-organize them in a summary according to the important events that the sentences describe. Events are defined by the event terms and the associated entities at the sentence level. In this paper, we emphasize on the event semantic relations derived from external linguistic resource. Firstly, the graph based on the event semantic relations is constructed and the events in the graph are grouped into clusters using the revised DBSCAN clustering algorithm. Then, we select one event as the representative event for each cluster or one cluster to present the main topic of the documents. Lastly, we generate the summary by extracting the sentences which contain more informative representative events from the documents. The evaluation on the DUC 2001 document sets shows it is necessary to take the semantic relations among the events into consideration and our summarization approach based on event semantic relation graph clustering is effective.
Original language | English |
---|---|
Pages (from-to) | 721-730 |
Number of pages | 10 |
Journal | International Journal of Advancements in Computing Technology |
Volume | 4 |
Issue number | 22 |
DOIs | |
Publication status | Published - 1 Dec 2012 |
Keywords
- DBSCAN clustering algorithm
- Event graph
- Event-based summarization
- VerbOcean
ASJC Scopus subject areas
- General Computer Science