摘要
[目的/意义]针对基于内容的个性化推荐策略,提出资源特征选择与权值计算优化策略,从而改善个性化推荐的效果。[方法/过程]构建基于用户决策机理的个性化推荐模型,模型以用户决策机理为背景知识进行资源特征的选择、用户兴趣模型的构建与语义表示、用户决策函数构建。为验证模型效果,以4 748位用户的观影数据为例进行实验,实验以向量空间模型为参照模型,P@N为评价指标。[结果/结论]实验结果显示,在N取值为5、10、20、50、100、200的情况下,基于用户决策机理的个性化推荐模型效果都显著优于向量空间模型,从而验证模型的有效性。
[Purpose/significance] The purpose of this paper is to propose an optimization strategy of features choosing and weight computing for content-based personalized recommendation.[Method/process] This paper proposes a personalized recommendation model based on user's decision-making mechanism,which takes user decision mechanism as background knowledge in features selection,user interest profile construction and semantic representation,and user decision function construction.To test this model,this paper conducts an experiment taking 4 748 users as sample,vector space model as reference model,and P@N as evaluation index.[Result/conclusion] The results show that,in the cases of N equals 5,10,20,50,100,200,the personalized recommendation model based on user decision-making mechanism is significantly better than the vector space model,and the effectiveness of the model is verified.
作者
林鑫
桑运鑫
龙存钰
Lin Xin;Sang Yunxin;Long Cunyu(Institute of Scientific and Technical Information of China, Beijing 100038;School of Information Management of Central Normal University, Wuhan 430079;School of Information Management of Wuhan University, Wuhan 430072)
出处
《图书情报工作》
CSSCI
北大核心
2019年第2期99-106,共8页
Library and Information Service
基金
国家社会科学基金青年项目"社会网络中基于用户认知结构的知识标注研究"(项目编号:17CTQ024)研究成果之一
关键词
决策机理
基于内容推荐
个性化推荐
decision-making mechanism
content-based recommendation
personalized recommendation