Approach for topical sentence extraction based on model LDA
Wang Li · Computer Engineering and Applications Journal · 2013
This paper proposes a novel topic sentence extraction approach based on model LDA on basis of acquiring candidate topic sentences through the topic-related and query-based keyword expansion.It can extract fine granularity on the subject and increase the reliability of the certain topical information.On several sub-topics against a target topic,it extracts those topic sentences by means of the reliability calculation according to the smoothness of the topic-sentence probability distribution.The method achieves good result in the special application of sentence extraction on web topic.