摘要
鉴于电子商务网站推荐系统的需要,将用户兴趣分为长期兴趣和短暂兴趣,并提出一种基于长期兴趣和短暂兴趣的用户偏好表示法.利用web服务器数据库的数据,采用无监督学习方法,对用户注册信息进行挖掘,提取出用户长期兴趣.基于向量映射,对web服务器日志上的用户使用记录数据和内容数据进行分析,提取用户短暂兴趣.通过用户反馈信息修正"粗糙"用户偏好文档,使得用户偏好文档更新得以实现.最后,应用了实证案例验证了该方法的合理性和有效性.
Abstract: In view of the needs of E-commerce website for recommendation system, user interests are divided into the long-term interest and the short-term interest, furthermore, based on the long-term interest and the short-term interest, a way to describe users' preference is proposed. On the basis of the data from the web server database, users' registration information can be fully mined to abstract users' long-term interest by using unsupervised learning. Both the records data and content data on the server log are analyzed to abstract users' short-term interest by vector mapping. Moreover, the rough profile presenting users' preference can be modified by dealing with users' feedback, as a rescut, updating users' preference profile becomes possible. Case analysis illustrates to a certain extent this method is reasonable and feasible.
出处
《同济大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2013年第6期953-960,共8页
Journal of Tongji University:Natural Science
基金
国家自然科学基金(70971099)
中央高校基本科研业务费专项资金(1200219198)
上海市科技发展基金软科学研究博士生学位论文资助(12692193000)
关键词
WEB数据挖掘
长期兴趣
短暂兴趣
用户偏好
Key words~ web data mining~ long-term interest~ short-terminterest~ users' preference
作者简介
王洪伟(1973-),男,副教授,博士生导师,管理学博士,主要研究方向为商务智能与情感计算.E-mail:hwwang@tongji.edu.cn