一个基于Agent的个性化推荐系统  

An Agent-Based Personalized Recommendation System

作　　者：胡迎松[1] 韩苹[1] 陈中新[2] 

机构地区：[1]华中科技大学计算机科学与技术学院,湖北武汉430074 [2]湖北清江水电开发有限责任公司,湖北宜昌443002

出　　处：《计算机应用研究》2006年第4期78-80,共3页Application Research of Computers

摘　　要：目前的推荐系统有基于文本过滤、基于协同过滤、基于W eb挖掘等多种类型,但都只适用于某个应用领域,难以满足推荐应用的多种推荐需求。引入网络智能的思想,利用多Agent(智能代理)技术提出了一个基于Agent的个性化推荐系统架构———ABPRS。该推荐系统能满足多种推荐需求,在用户看来网络具有一定智能。There have been various approaches for recommendation systems, such as content-based, collaborative, Web mining-based and so on. The performance of all these recommendation methods varies according to the characteristic of the application domain to which they are applied. It is difficult for any one of these methods to satisfy multiple recommendation requests. This paper presents a personalized recommendation system with Multi-Agents based on Web intelligence thought. This recommendation system can satisfy multiple recommendation requests and the Internet will appear a little intelligent in the view of users.

关 键 词：个性化 推荐系统 网络智能 智能代理 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]