基于社会化标注的个性化推荐研究进展

Advances in Personalized Information Recommendation Based on Social Tagging

社会化标注是当前互联网研究中的一个热点。本文在对社会化标注的内涵和结构加以简单介绍的基础上,重点探讨了基于社会化标注进行推荐的相关进展。首先是明确了标签对于用户模型的意义,接着,从用户、资源和标签三个角度对基于社会化标注的聚类算法进行了讨论。同时也对基于社会化标注的排序算法进行了分析,并进一步将其分为依附补充、独立排序和通用排序三类算法。然后,对标签推荐方面的研究进行了探讨,主要是围绕内容分析、协同分析、语义分析三个方面展开的。最后,分析了社会化标注中个性化信息推荐的研究,发现借助矩阵、聚类和网络的分析是三种主要思路。

Social tagging is a hot topic in current Internet-related researches.Based on the introduction of the meaning and structure of social tagging,this paper mainly discusses the advancements of personalized information recommendation based on social tagging.Firstly,tag＇ s meaning for user profile is proved,and clustering algorithms based on social tagging from the aspects of users,resources and tags are discussed.Meanwhile,ranking algorithms based on social tagging are also studied,and find supplementary,independent and universal ranking are three sub-algorithms.Afterwards,researches about tag recommendation are discussed,which mainly focus on means of content,collaborative and semantic analysis.Finally,studies on personalized information recommendation based on social tagging are analyzed,and find matrix,clustering and network analysis are three primarily methods.