摘要
全面探讨了电子商务个性化推荐系统,对推荐系统所使用的多种智能化的推荐技术进行了分类讨论,分析归纳了各种推荐技术的优点与不足之处,在现有的混合推荐技术上提出了一种基于用户兴趣聚类的协同过滤推荐改进技术,然后对改进的推荐技术的有效性进行了实验验证,结果表明基于用户兴趣聚类的推荐技术有效解决了用户评分较少而造成推荐困难的问题,显示出比传统的推荐方法更好的推荐质量和扩展性.
In the paper personalized e-commerce recommendation system was firstly discussed from the point of view as representation way.Secondly,a variety of intelligent recommendation techniques in recommendation system were analyzed to summarize the advantages and disadvantages of every recommendation techniques,and then put forwards a collaborative filtering recommendation improved technique based on users interest clustering.The effectiveness of recommendation improved technique was validated by experiments,and the results showed that the collaborative filtering recommendation improved technique based on users interest clustering effectively solved the difficult recommendation because of less user rating and gave a better recommendation quality and scalability than the traditional recommendation methods.
出处
《商丘职业技术学院学报》
2011年第2期41-44,共4页
JOURNAL OF SHANGQIU POLYTECHNIC
基金
2010年河南省政府决策研究招标课题(编号:B563)
关键词
电子商务
智能推荐技术
协同过滤
E-commerce
Intelligent recommended technology
collaborative filtering
