摘要
微博话题推荐算法的作用是当用户面临微博信息过载时,结合用户的基本信息,帮助用户找到对自己有价值的微博话题。微博推荐算法的核心任务是以用户信息为基础,分析用户的偏好,并推荐给其他信息相似的用户。本文提出的基于用户聚类的微博推荐算法包括三个层次,即用户微博话题特征提取、用户聚类、微博话题推荐。实验表明该系统的准确率达到50.2%,可准确地为用户进行微博话题推荐,并提高了用户浏览微博的效率。
The recommendation algorithm of micro-blog topic aims to help users find valuable micro-blog topic based on the users＇ basic information when overload information of micro-blog has occurred. The main tasks of the micro-blog recommendation algorithm are analyzing the users＇ preferences and recommending a special micro-blog topic to other users with similar information. This paper proposes a user clustering-based micro-blog topic recommendation algorithm which includes three levels, namely the users micro-blog topic features extraction, users clustering, and users recommendation. Experimental results show that the accuracy of the proposed system is up to 50.2%. It can accurately recommend micro-blog topic for users. Thus, the efficiency of browsing the micro-blog can be improved greatly.
出处
《阜阳师范学院学报（自然科学版）》
2016年第2期74-79,共6页
Journal of Fuyang Normal University(Natural Science)
基金
安徽省教育厅自然科学基金重点项目(KJ2015A111)
上海市信息安全综合管理技术研究重点实验室(上海交通大学)开放课题(AGK2013002)资助
关键词
微博话题
用户聚类
推荐
Micro-blog topic
user clustering
recommendation
