Dynamic ontology-based user modeling in personalized information retrieval system
Personalized retrieval is the inevitable trend of information services, which can meet users diversified and specialized knowledge requirements. The learning of user features and the building of user model are essential to realize the personalization in information retrieval system. In this paper, we introduced ontology into this research, and study the representation of user interests using concepts and relations, and based on which, develop a dynamic approach of modeling. We analyzed the main principle and the two key steps of this approach including user features primitive learning and deep mining. The application of the model is also discussed at the end of the paper. Our study shows that dynamic ontology-based user modeling can improve the retrieval quality by better representing, discovering and utilizing user implicit interests.
ontology user modeling information retrieval personalization
Dangxiang Ai Hui Zuo Gaoyong Liu
School of Management, Guangdong University of Technology, Guangzhou, China School of Economics and Tade, Guangdong University of Technology, Guangzhou, China
国际会议
北京
英文
139-143
2010-11-27（万方平台首次上网日期，不代表论文的发表时间）