TAG RECOMMENDATION BASED ON TAG-TOPIC MODEL

摘要：

　　With the rapid increase of the social websites,most social tagging systems are allowing users to share and to label various kinds of resources with their favorite tags.However,the uncontrolled use of tags makes the resources attached with some irrelevant even noise tags.To solve the problem,this paper proposes a tag-topic model based approach to recommend tags for resources,which elicits latent topics from resources and maps new resources to these latent topics so as to recommend the most appropriate tags for the resources.The experimental results show the effectiveness of the proposed approach.

关键词： Tag recommendation Tag-topic model

作者: Rong Hu Tingting He Fang Li Po Hu

作者单位: School of Computer Science,Central China Normal University,Wuhan 430079,China National Engineering Research Center for E-Learning,Central China Normal University,Wuhan 430079,Chi

会议类型: 国际会议

会议名称: 2012 2nd IEEE International Conference on Cloud Computing and Intelligence Systems (2012年第2届IEEE云计算与智能系统国际会议(IEEE CCIS2012))

会议地点: 杭州

会议语种:英文

页码: 1983-1987

在线出版日期: 2012-10-30（万方平台首次上网日期，不代表论文的发表时间）