结合超图投影和随机游走的个性化推荐方法  

An Approach for Personalized Recommendation via Hypergraph Projection and Random Walk

作　　者：苏晓萍[1] 楼俊钢[1] 

机构地区：[1]湖州师范学院信息与工程学院,湖州313000

出　　处：《情报学报》2012年第8期830-836,共7页Journal of the China Society for Scientific and Technical Information

基　　金：国家自然科学基金：基于核函数回归估计理论的软件可靠性建模研究（61103051）;国家自然科学基金：基于反馈的三维交互式动态GIS城市规划模型研究（60872057）;湖州市自然科学基金：无线传感器网络上的传播动力学研究（2011YE08）.

摘　　要：以大众分类（folksonomy）为基础的社会化网络中，通常用标签对资源特征和用户喜好进行描述，本文将标签语义引人传统的推荐系统，从用户的标注行为入手，提出一种基于超图投影的推荐方法，该方法利用投影图中节点的连边权重进行节点相似性度量，使待推荐对象在投影图上随机游走，根据待推荐对象在节点上停留概率获得推荐，实验结果表明标签的引入提高了推荐质量，算法在精确性和多样性上均有很好的改进。In folksonomy-based social networks, tags are important information which can be used to describe features of recourses as well as users＇ topic preferences. The latent semantics of Tags are introduced into traditional recommendation system, then starting from user＇ s annotation an approach for recommendation was proposed based on hypergraph projection. In our approach, the similarity of nodes is measured by edge weights in projected hypergraph, the objects to be recommended random walks on the projected hypergraph, so that the result of recommendation can be obtained according to its＇ probability of staying on the node. The experimental results clearly show that our approach is effective, both diversity and accuracy can be improved by considering the latent semantics of Tags.

关 键 词：大众分类 标签 个性化推荐 超图 随机游走 

分 类 号：F832.5[经济管理—金融学]