摘要
随着移动互联网的进步和信息量的急剧增长,信息过载使得用户获取需求信息更加困难。由于推荐系统可以较好地解决信息过载问题,因而被广泛应用于各种移动网络平台。在推荐系统中,应用最为广泛和成功的一种技术是协同过滤推荐。本文首先介绍了协同过滤推荐技术的原理、分类和存在的问题,然后简要概括了评价推荐系统是比较常用的评估方法,并对进一步需要研究的问题进行总结。
With the progress of mobile Internet and the rapid growth of information, information overload makes itmore difficult for users to obtain information. As the recommendation system can solve the problem of informationoverload well, it is widely used in various mobile network platforms. One of the most widely used and successful appli-cations in recommender systems is collaborative filtering recommendation. This paper first introduced the principle,classification and existing problems of collaborative filtering recommendation technology, and then briefly summa-rized that the evaluation and recommendation system was a commonly used evaluation method, and summarized thefurther research problems.
引文
[1]李勇.移动互联网信息安全威胁与漏洞分析[J].通信技术,2014(4):439-444.
[2]Ricci F.Mobile recommender systems.[J].InformationTechnology&Tourism,2010(6):205-231.
[3]张忠平,郭献丽.PEV:一种新的用于Item—Based协同过滤算法的相似性度量方法[J].小型微型计算机系统,2009(4):716-720.
[4]彭玉,程小平.基于属性相似性的Item-based协同过滤算法[J].计算机工程与应用,2007(14):144-147.
[5]董立岩,王越群,贺嘉楠,等.基于时间衰减的协同过滤推荐算法[J].吉林大学学报(工学版),2017(4):1268-1272.
[6]Davidsson C,Moritz S.Utilizing implicit feedback and con-text to recommend mobile applications from first use[C]//The Workshop on Context-Awareness in Retrieval&Recommendation.ACM,2011.