摘要
随着网络技术的发展、网络上信息量的不断增多以及电子商务的普及,协作过滤得到了广泛的应用。在对协作过滤信息推荐技术基本思想、算法分类等相关内容概述的基础上,总结了其当前存在的3个方面的主要问题:冷启动、数据高维稀疏、以及算法的可伸缩性,同时分析了当前国际、国内的研究进展情况,对主要的方法、技术等进行了扼要的概述与总结。对推荐算法中相似性度量方法、预测评分的策略、推荐质量的度量标准、未来可能的研究方向等分别进行了论述。
With the development of network technology and E-commerce,and the increasing information,collaborative filtering has become the most successful technology of information recommendation.This paper presents an overview of collaborative filtering recommender systems,summarizes the three main issues that currently exist,analyzes the main techniques applied in the current generation of collaborative filtering algorithm,describes various limitations of current recommendation methods,and finally discusses the possible ways and possible future research directions that can improve the recommendation quality.
出处
《重庆邮电大学学报(自然科学版)》
北大核心
2012年第1期78-82,共5页
Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)
基金
重庆市教委项目(10SKF10)~~
关键词
个性化服务
协作过滤
稀疏性
伸缩性
personalized service
collaborative filtering
scarcity
scalability