基于Vague值的非个性化产品推荐研究  

Study of non-personalized recommender systems based on vague value

作　　者：崔春生[1] 苏白云[2] 

机构地区：[1]河南财经政法大学计算机与信息工程学院,郑州450002 [2]河南财经政法大学数学与信息科学系,郑州450002

出　　处：《计算机工程与应用》2012年第13期63-66,共4页Computer Engineering and Applications

基　　金：河南省科技厅基础与前沿课题;河南省教育厅基础与前沿课题

摘　　要：在网络用户和网络产品急剧攀升的背景下,非个性化产品推荐成为一种很好的网络广告手段,已有的研究中,Vague集方法已被应用于推荐系统中,并取得了较好的效果。分析了非个性化产品推荐的一般特征和优点;借助Vague值描述的产品,研究了特征值方法和"马太效应"记分函数方法运用于产品排序的可行性;最后,通过实例验证了两种方法的在非个性化产品推荐中的有效性和一致性。Non-personalized recommender systems has become a popular online advertising method due to the comprehensive applications of users and products. In the previous research, Vague value method has been used to study the recommender system, which is approved to be used in this field. The general characteristics and advantages of non-personalized recommender are reviewed. Hence, the feasibilities of using eigenvalue method and the ＂Matthew Effect＂ score function method in products-ranking are studied based on description of products by Vague set. The effectiveness and coconscious of two methods are then proved with examples.

关 键 词：非个性化推荐 VAGUE值 特征值 “马太效应”函数 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]