摘要
文中给出了一种新的数据源的获取方法,使用Web2.0技术直接从客户浏览行为中获取需要的数据,避免了传统Web使用数据挖掘时日志数据预处理时的大量繁杂工作,减少了噪声数据,提高了数据准确性。根据所获数据建立用户-商品矩阵,计算此矩阵的欧氏距离,在此基础上使用聚类算法将客户进行聚类,根据聚类结果对新来的客户进行有目的的商品推荐,并对聚类结果进行跟踪评价。目的是为了提高电子商务网站的个性化服务。
In this paper, proposed a new approach to achieve the data in order to cluster the customers; use the Web2.0 technology to acquire information from the behavior of customers who visited the Web site. Therefore, can avoid the lots of heavy work of the pretreatment of Web log data. Then, build the customer- commodity matrix and figure out the distance of Euclidean, and cluster the customers by using the arithmetic that is based on commodity recommendation of clustering. Based on the clustering result, recommend commoditiea to the new customers in order to see their reactions, and evaluate the result of customers clustering.
出处
《计算机技术与发展》
2008年第7期212-214,221,共4页
Computer Technology and Development
基金
辽宁省教育科学研究项目(05L338)
关键词
电子商务Web使用挖掘
商品推荐
个性化
客户聚类
electronic commerce
web use mining
commodity recommendation
personalization
customers clustering
