计算机应用 ›› 2015, Vol. 35 ›› Issue (5): 1324-1327.DOI: 10.11772/j.issn.1001-9081.2015.05.1324

• 人工智能 • 上一篇    下一篇

在推荐系统中利用时间因素的方法

范家兵1,2, 王鹏3, 周渭博1,2, 燕京京1,2   

  1. 1. 中国科学院 成都计算机应用研究所, 成都 610041;
    2. 中国科学院大学, 北京 100049;
    3. 成都信息工程学院 软件工程学院, 成都 610225
  • 收稿日期:2014-12-02 修回日期:2015-01-13 出版日期:2015-05-10 发布日期:2015-05-14
  • 通讯作者: 范家兵
  • 作者简介:范家兵(1990-),男,安徽六安人,硕士研究生,主要研究方向:个性化推荐; 王鹏(1975-),男,四川犍为人,教授,博士,CCF会员,主要研究方向:云计算、并行计算、量子算法; 周渭博(1981-),男,山东烟台人,博士研究生,主要研究方向:大数据智能处理; 燕京京(1989-),女,河南安阳人,硕士研究生,主要研究方向:数据挖掘.
  • 基金资助:

    国家自然科学基金资助项目(60702075);广东省科技厅高新技术产业化科技攻关项目(2011B010200007);四川省青年科学基金资助项目(09ZQ026-068);成都市科技局创新发展战略研究项目(11RXYB016ZF);四川省科技创新苗子工程项目(2014-063).

Method by using time factors in recommender system

FAN Jiabing1,2, WANG Peng3, ZHOU Weibo1,2, YAN Jingjing1,2   

  1. 1. Chengdu Institute of Computer Application, Chinese Academy of Sciences, Chengdu Sichuan 610041, China;
    2. University of Chinese Academy of Sciences, Beijing 100049, China;
    3. College of Software Engineering, Chengdu University of Information Technology, Chengdu Sichuan 610225, China
  • Received:2014-12-02 Revised:2015-01-13 Online:2015-05-10 Published:2015-05-14

摘要:

针对传统推荐算法忽略时间因素的问题,根据个体用户短期行为的相似性,利用时间衰减函数计算项目间相关关系,提出基于用户兴趣的项目关联度; 将其用于项目相似度的计算,提出基于用户兴趣的项目相似度; 同时基于项目关联度对ItemRank算法进行改进,提出一种结合时间因素的TItemRank算法.实验结果表明, 利用项目关联度对推荐算法进行改进时,在推荐项目数较少的情况下能够明显地改善推荐效果.特别地,在推荐项目数为20时,基于用户兴趣的项目相似度相比余弦相似度和Jaccard相似度,推荐准确率分别提高了21.9%、6.7%; 在推荐项目数为5时,TItemRank算法相比ItemRank算法推荐准确率提高2.9%.

关键词: 协同过滤, 项目关联度, 项目相似度, 兴趣衰减, ItemRank, 图模型, 艾宾浩斯曲线

Abstract:

Concerning the problem that traditional recommendation algorithm ignores the time factors, according to the similarity of individuals' short-term behavior, a calculation method of item correlation by using time decay function based on users' interest was proposed. And based on this method, a new item similarity was proposed. At the same time, the TItemRank algorithm was proposed which is an improved ItemRank algorithm by combining with the user interest-based item correlation. The experimental results show that: the improved algorithms have better recommendation effects than classical ones when the recommendation list is small. Especially, when the recommendation list has 20 items, the precision of user interest-based item similarity is 21.9% higher than Cosin similarity and 6.7% higher than Jaccard similarity. Meanwhile, when the recommendation list has 5 items, the precision of TItemRank is 2.9% higher than ItemRank.

Key words: collaborative filtering, item correlation, item similarity, interest in attenuation, ItemRank, graph model, Ebbinghaus curve

中图分类号: