基于协作反馈的蚁群算法的自适应网站研究  

A study on adaptive website based on ant colony algorithm with multi-agent cooperation and feedback mechanism

作　　者：黄艳欢[1] 何振峰[1] 

机构地区：[1]福州大学数学与计算机科学学院,福建福州350002

出　　处：《福州大学学报（自然科学版）》2008年第6期814-818,共5页Journal of Fuzhou University(Natural Science Edition)

基　　金：福建省青年创新基金资助项目(2006F3075)

摘　　要：提出了一种基于蚁群算法的关联网页推荐方式,针对蚁群算法中固有的缺陷以及它在自适应网站应用中所出现的问题进行相应的模型改进.引入了选择权值自适应,反馈机制和多智能体协作的思想,解决了收敛速度和停滞现象的矛盾平衡问题,实现了动态调整系统推荐和不同访问用户之间的交互,并根据用户日志进行了访问倾向性推荐.This paper proposes an approach to recommend related web pages based on the ant colony algorithm, and makes some improvements on the model concerning to the problems existing in it and the ones arising in the application to the adaptive website. The paper mainly introduces the adaptive weight of the decision, feedback mechanism and multi - agent cooperation, and keeps the balance be- tween the convergence rate and the stagnancy, realizes the dynamic adjustment of the system recommendation and the communication of different visitors, logs by the users. finally makes recommendation according to the

关 键 词：蚁群算法 自适应网站 反馈机制 多协作智能体 

分 类 号：TP393.09[自动化与计算机技术—计算机应用技术]