虚拟社区中基于相似标签聚类的语义信息推荐  

Semantic Information Recommendation Based on Similar Tags Clustering in Virtual Community

作　　者：周朴雄[1] 陈涛[1] 

机构地区：[1]华南理工大学经济与贸易学院,广东广州510006

出　　处：《情报理论与实践》2013年第10期100-104,共5页Information Studies:Theory & Application

摘　　要：自然语言的标注方式导致标签产生了模糊性和多样性问题。由用户、资源和标签三者关系共同确定了语义本体,即确定了标签的唯一性,同时也建立了标签之间的语义联系。在此基础上,文章提出一种应用在虚拟社区中,基于相似标签聚类的信息推荐模型。该模型通过对标签间的相似性进行计算、聚类,以聚类的标签群作为纽带,形成信息资源的语义链条,进而挖掘出用户的需求信息,最后以标签云的方式展示给用户,完成信息推荐过程。The labeling method of natural language has resulted in the ambiguity and diversity problems of tags. The relationships among users, resources and tags determine the semantic ontology together, that is, determine the uniqueness of tags, and at the same time establish the semantic relationships between tags. On this basis, this paper proposes an information recommendation model based on similar tags clustering, which can be used in the virtual community. The model calculates and clusters the similarities of tags, takes the clustered tag group as the link to form the semantic chains of information resources, and excavates the information needed by users. Finally, the information is brought forth to the users in the way of tag cloud, thus completing the information recommendation process.

关 键 词：虚拟社区 标签聚类 信息推荐 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]