摘要
针对用户个人特征并向其提供准确恰当信息的个性化信息推荐研究,一直是学术界和产业界所关注的热点。结合后控词表,对用户分散的、个性化的标注进行处理,并将用户兴趣用向量表示,然后借鉴协同过滤算法的思想,寻找出相似用户集及其内部的资源集。在此基础上,采用相对匹配策略,提出一种基于社会化标签系统的个性化推荐方法。
Research on how to provide accurate and appropriate recommendations based on user profile has always been a hot spot. Combining the handling of controlled vocabulary, this paper seeks to deal with users＇ discrete and personalized tags. Firstly, it expresses the user interest as a user interests vector, then finds similar users and the resources, and draws on the collaborative fihering algorithms. On this basis, by consulting the relative matching method, the paper presents a personalized recommendation method which based on the social tagging system.
出处
《图书情报工作》
CSSCI
北大核心
2010年第1期50-53,120,共5页
Library and Information Service
关键词
社会化标签系统
个性化信息推荐
协同过滤
social tagging system personalized information recommendation cooperative filtering
