• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2016, Vol. 38 ›› Issue (04): 761-767.

• 论文 • 上一篇    下一篇

基于多项式有限混合模型的Slope One算法改进

陈平华1,王旭彬1,洪英汉2   

  1. (1.广东工业大学计算机学院,广东 广州 510006;2.韩山师范学院,广东 潮州 521000)
  • 收稿日期:2015-01-29 修回日期:2015-07-10 出版日期:2016-04-25 发布日期:2016-04-25
  • 基金资助:

    广东省教育部产学研结合项目(2012B091100003,2012B091000058);广东省专业镇中小微企业服务平台建设项目(2012B040500034)

An improved Slope One algorithm based on
multinomial finite mixture model           

CHENG Pinghua1,WANG Xubin1,HONG Yinghan2   

  1. (1.School of Computers,Guangdong University of Technology,Guangzhou 510006;
    2.Hanshan Normal University,Chaozhou 521000,China)
  • Received:2015-01-29 Revised:2015-07-10 Online:2016-04-25 Published:2016-04-25

摘要:

针对多项式有限混合模型参数估计过程中存在的初始化依赖、参数易收敛到边界值以及容易陷入局部最优等问题,引入了最小信息长度准则,优化多项式有限混合模型的参数估计过程。在此基础上,采用基于多项式有限混合模型的聚类算法对用户评分行为进行聚类,利用模型求解得到的聚类归属概率对Slope One算法实施改进。实验结果表明:应用最小信息长度准则对多项式有限混合模型进行优化后,聚类效果明显提高;同时,相比于基于用户聚类的Slope One推荐算法,改进算法具有明显的改进效果。

关键词: 有限混合模型, Slope One, 基于模型聚类, 协同过滤

Abstract:

The problems of the multinomial finite mixture model such as initialization dependence, the boundary of the parameter space and the local optimum are solved with the minimal message length criterion in this paper. We propose a multinomial finite mixture model based algorithm and apply it to clustering the behavior data of user ratings. Besides, the probability of class belonging is used to improve the Slope One algorithm. Experimental results prove that the new clustering algorithm does a better job on clustering behavior data and the new recommendation algorithm outperforms several other recommendation algorithms.   

Key words: finite mixture model;Slope One;modelbased clustering;collaborative filtering