标题:Based on the Chinese question relevant search algorithm research within the restricted areas
作者:Li, Z.X. ;Han, H.J.
作者机构:[Li, Z.X. ;Han, H.J. ] School of Computer Science and Technology, Shandong University of Finance and Economics, Jinan; Shandong, China;[Li, Z.X. ;Han, 更多
会议名称:International Conference on Multimedia, Communication and Computing Application, MCCA 2014
会议日期:October 15, 2014 - October 16, 2014
来源:Multimedia, Communication and Computing Application - Proceedings of the International Conference on Multimedia, Communication and Computing Application, MCCA 2014
出版年:2015
页码:483-491
摘要:This paper presents a kind of search algorithm based on the Chinese question correlation. We determine the longitudinal correlation between the words through creating a keyword tree. We establish a longitudinal correlation matrix, and get the best answer to the query by using the correlation calculation. This method firstly extracts the semantic representation of a question sentence through “question sentence semantic representation extracting model” which extends the semantic representation of the question sentence, and then we search the answer of the question sentence by using the area keywords tree. This method adopts a strategy of keywords replacement during the process of further search. Once the keywords replacement has completed, the answer of the question sentence will be searched. This method can be integrated into an intelligent question answering system easily. Experimental results suggest that this method is effective.
© 2015 Taylor & Francis Group, London.
收录类别:EI
资源类型:会议论文
TOP