A SEMI-SUPERVISED LEARNING METHOD COMBINED WITH DIMENSIONALITY REDUCTION IN VIETNAMESE TEXT SUMMARIZATION

摘要

The World Wide Web has brought us a vast amount of online information. When we search with a keyword, we get data feedback from many different websites and the user cannot read all the information. So, text summarization has become a hot topic, and it has attracted experts in data mining and natural language processing field. For Vietnamese, some methods of text summarization that have been proposed for English also bring some significant results. However, there still remain some difficult problems to treat with the Vietnamese language processing; typical in this are the Vietnamese text segmentation tool and text summarization corpus. In this paper, we present a Vietnamese text summarization method based on sentence extraction approach using neural network for learning combined with reducing dimensional features to overcome the cost when building term sets and reduce the computational complexity. The experimental results show that our method is really effective in reducing computational complexity, and is better than some methods that have been proposed previously.