计算机应用 ›› 2016, Vol. 36 ›› Issue (10): 2789-2793.DOI: 10.11772/j.issn.1001-9081.2016.10.2789

• 人工智能 • 上一篇    下一篇

融合标签与人工蜂群的微博推荐算法

王宁宁1,2, 鲁燃1,2, 王智昊1,2   

  1. 1. 山东师范大学 信息科学与工程学院, 济南 250014;
    2. 山东省分布式计算机软件新技术重点实验室, 济南 250014
  • 收稿日期:2016-04-21 修回日期:2016-06-24 出版日期:2016-10-10 发布日期:2016-10-10
  • 通讯作者: 鲁燃,E-mail:wangningning_5678@163.com
  • 作者简介:王宁宁(1991—),女,山东德州人,硕士研究生,CCF会员,主要研究方向:话题追踪、热点发现、微博推荐;鲁燃(1972—),男,山东郓城人,副教授,主要研究方向:网络信息安全;王智昊(1986—),男,山东济南人,博士研究生,主要研究方向:信息安全。
  • 基金资助:
    国家自然科学基金资助项目(61373148,61502151);国家社会科学基金资助项目(12BXW040);山东省自然科学基金资助项目(ZR2012FM038,ZR2014FL010);山东省优秀中青年科学家奖励基金资助项目(BS2013DX033);教育部人文社会科学基金资助项目(14YJC860042);山东省社会科学规划项目(15CXWJ13);山东省高等学校科技计划项目(J15WB37,J15LN02)。

Micro-blog recommendation algorithm by combining tag and artificial bee colony

WANG Ningning1,2, LU Ran1,2, WANG Zhihao1,2   

  1. 1. College of Information Science and Engineering, Shandong Normal University, Jinan Shandong 250014, China;
    2. Shandong Provincial Key Laboratory for Distributed Computer Software Novel Technology, Jinan Shandong 250014, China
  • Received:2016-04-21 Revised:2016-06-24 Online:2016-10-10 Published:2016-10-10
  • Supported by:
    BackgroundThis work is partially supported by National Natural Science Foundation of China (61373148,61502151), the National Social Science Foundation of China (12BXW040), the Natural Science Foundation of Shandong Province (ZR2012FM038,ZR2014FL010), the Outstanding Young Scientist Award Foundation of Shandong Province (BS2013DX033), the Social Science Foundation of Ministry of Education of China (14YJC860042), the Social Science Project of Shandong Province (15CXWJ13), the Higher Educational Science and Technology Program of Shandong Province (J15WB37,J15LN02).

摘要: 针对基于标签的推荐算法中存在的冷启动问题,提出了一种融合标签与人工蜂群的微博推荐算法——TABC-R。首先,对用户的标签信息进行定义,并使用标签集表示用户兴趣;其次,根据标签权重、标签属性权重和标签与微博中词语的相似度三种变量来构造人工蜂群算法中的适应度函数;最后,利用人工蜂群算法的搜索策略,搜索出具有最优适应度值的微博向用户进行推荐。与基于标签的推荐(T-R)算法和基于人工蜂群的推荐算法(ABC-R)相比,TABC-R算法的准确率和召回率均有小幅提升,表明了TABC-R算法的有效性。

关键词: 标签, 人工蜂群, 微博, 冷启动, 适应度函数

Abstract: Focusing on the cold-start problem existed in the recommendation algorithm, a micro-blog Recommendation algorithm combined Tag and Artificial Bee Colony, namely TABC-R, was proposed. Firstly, the tag information for user was defined, and the tag set was used as user's interest. Secondly, the fitness function of Artificial Bee Colony (ABC) algorithm was established by three variables including tag weight, tag attribute weight and the similarity of the micro-blog words and the tags. Finally, the micro-blog with the best fitness value was obtained and recommended to users according to the search strategy of ABC algorithm. Compared with Tag-based Recommendation (T-R) algorithm and the Recommendation algorithm based on ABC (ABC-R), TABC-R algorithm has a light increase in the precision and recall, which proves the effectiveness of TARC-R.

Key words: tag, Artificial Bee Colony (ABC), micro-blog, cold-start, fitness function

中图分类号: