计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (15): 163-165.

• 数据库与信息处理 • 上一篇    下一篇

基于蚁群计算的自适应Web检索算法设计

陶剑文   

  1. 浙江工商职业技术学院 计算机应用研究所,浙江 宁波 315012
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-05-21 发布日期:2007-05-21
  • 通讯作者: 陶剑文

Algorithm design for adaptive Web retrieval system based on ant system computation

TAO Jian-wen   

  1. Computer Application Institute,Zhejiang Business Technology Institute,Ningbo,Zhejiang 315012,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-05-21 Published:2007-05-21
  • Contact: TAO Jian-wen

摘要: 受蚁群觅食行为仿生研究和蚁群系统模型理论所启发,提出了一种基于蚁群计算模型的分布、协作多主体(multi-agent)反应架构的自适应、可伸缩的Web搜索系统模型(MASAIR),其由大量智能主体组成,利用智能主体架构的优异特性,旨在从巨型超文档集合(Web)中自治地搜索特定主题的信息,从而为用户提供迅捷的信息检索服务。详细描述了MASAIR的计算模型及其算法,通过对标准Web文档集的检索仿真实验结果显示:该架构具有对环境改变的鲁棒性和对用户信息需求变更的自适应性。

关键词: 智能主体, 信息检索, 相似性测量, 多主体系统

Abstract: This paper presents an adaptive and scalable Web search system,based on a multi-agent reactive architecture,which draws inspiration from biological researches on the ant foraging behavior.Its target is to search autonomous information on particular topics,in huge hyper textual collections,such as the Web,exploiting the outstanding properties of the agent architectures.The algorithm has been proven to be robust against environmental alterations and adaptive to user’s information need changes,discovering valuable evaluation results from standard Web collections.

Key words: intelligent agent, information retrieval, similarity measure, multi-agent system