摘要
个性化是未来Web智能系统的一大特征.为了实现商品的个性化推荐,提出了一种新的基于多级客户模型的推荐系统机制,它由数据准备、模型学习、推荐集的生成和智能过滤四个子过程构成.该机制借助于多级客户模型从客户的购物需求、偏爱特征和消费能力三方面捕获客户的实际需求,从而实现了一种深层次的个性化推荐,改善了推荐效果.
Personalization is a major characteristic of Web intelligent system for the future. To provide personalized commodity recommendation, this paper presents a new recommendation mechanism based on multilevel customer model, which consists of data preprocessing, model learning, recommendation set generating, and intelligent filtering. The shopping model of customer is learned from customer＇s shopping transactions firstly. By means of probability inference, the recommendation engine generates respective recommendation set for each online customer. Then based on preference and consumption model of customer, the commodities in the recommendation set are customized and orientated, which realizes more personalized commodity recom- mendation. In our approach, the system captures customer＇s practice needs from three aspects： shopping demands, preference characters and purchasing power, thus strengthens the degree of personalization in recommendation and improves the effectiveness of recommendation.
出处
《小型微型计算机系统》
CSCD
北大核心
2005年第9期1669-1673,共5页
Journal of Chinese Computer Systems
基金
国家自然科学基金项目(60173014)资助
北京市自然科学基金(4022003)资助
多媒体与智能软件技术北京市重点实验室开放基金(0702200201)资助.
关键词
个性化推荐
多级客户模型
贝叶斯网
推荐引擎
智能过滤
personalization recommendation
multilevel customer model
Bayesian network
recommendation engine
intelligent filter
