A knowledge induced graph-theoretical model for extract and abstract single document summarization
CICLing'13 Proceedings of the 14th international conference on Computational Linguistics and Intelligent Text Processing - Volume 2
Hi-index | 0.00 |
Single document summarization, which is as important as multiple document summarization for a variety of reasons, has been attracting declining interest recently. The goal of this study is to introduce a new approach to single document summarization and its implementation, SynSem. Our approach fuses syntactic, semantic, and statistical methodologies and reflects the importance of text headings in articles along with the presence of thematic keywords in sentences. Successful summary evaluation results are demonstrated when SynSem is tested on the Document Understanding Conference (DUC) 2002 data set using ROUGE, which compares single document summaries to baselines.