基于标签主题和概念空间的个性化推荐研究

Research on the Personalized Recommendation Based on Tag Topic and Concept Space

社会化标签已经成为个性化信息推荐领域的研究热点之一。为了克服标签推荐存在的不足,文章提出一种综合考虑标签主题和主题概念空间两种因素的标签推荐方法 (LDA-Concept)。通过主题下标签的推荐可以保证推荐的准确性,标签概念空间的推荐可以保证多样性。以MovieLens为平台进行实验,结果表明主题因素和概念空间因素有着同等的重要性,提出的LDA-Concept方法优于单纯使用LDA方法。

Social tag becomes one of the hot research issues on personalized information recommendation. In order to overcome the shortcomings of tag recommendation, this paper proposes a tag recommendation method （LDA-Concept） that comprehensively considers the factors of tag topic and topic concept space. Recommendation with tag topics can guarantee the accuracy of recommendation, and the recommendation of tag concept space can guarantee diversity. Taking MovieLens as the experimental platform, the experimental result shows that topic factor and concept space factor play equal importance. The LDA-Concept method proposed in the paper is superior to the LDA method.