摘要
分析了传统的基于项的协同过滤推荐算法在陶瓷行业电子商务平台应用中碰到的问题,提出了一种改进的基于产品特征的协同过滤推荐方法,建立陶瓷产品的特征矩阵,综合考虑产品特性和用户评价。应用表明能产生准确度较高的推荐结果,并能解决初期数据稀疏情况下的产品推荐问题。
Based on an analysis of the problems of the recommendation system in ceramic industry e-commeree platform, proposing an improved item-based collaborative filtering method based on product attribute,generating product attributes matrix, takeing both into account the influence of item attributes and user ratings. Applying results indicate these can provide better recommendation result and satisfactorily solve the problem when the user ratings are very sparse.
出处
《微计算机信息》
2009年第18期148-149,共2页
Control & Automation
关键词
电子商务
协同过滤推荐
产品特征
用户评价
陶瓷产品
E-Commerce
collaborative filtering recommendation
product attributes
customer commentation
ceramic product
