层次凝聚算法在商品个性化推荐中的应用

商品的个性化推荐是电子商务个性化服务中非常重要的一个方面,而聚类协作过滤则是推荐系统中采用最为广泛的技术。在基于聚类协作过滤的商品个性化推荐中的聚类算法通常采用划分聚类,文章根据电子商务网站的特点,提出了用改进的Rock层次凝聚算法Improved-Rock实现基于购买商品类别相似性的用户聚类。模拟实验结果表明该算法的应用是有实际价值的。

Applying Hiberarchy Clustering Arithmetic to Personalized Item Recommendation

Personalized item recommendation is one of the importantl aspects of personalization services given by e-business;meanwhile,collaborative Filtering technology based on clustering is the one most widely applied to recommendation syetem.Hiberarchy clustering arithmetic is the one often used in this system.In accordance with the characteristic of e-business webs,this paper forwards Improved-Rock arithmetic,an adaptive methods of belonged to hiberarchy clustering arithmetic.This arithmetic is appropriate to clustering based on similar sorts of item purchased by users.Simulative experiments indicate that the application of this arithmetic is practical and valuable.