Abstract: Since logistics resources on Internet are distributed and autonomous, their structures and interface are different. To mine global information in logistics resource-grid environment, a new grid-oriented data-mining model is provided without the same structure between grid resources. It is based on grid, agent and incremental data mining technology. Web-based logistics resource grid system is analyzed and resource-grid system is divided into logistics domains for easy management. Dynamic resource mining algorithm for a domain and request-based mining algorithm among domains are designed to resolve the data-mining problem among heterogeneous systems. Mining efficiency among domains is promoted with incremental data mining technology.

Key words: Resource grid, Incremental data mining, Logistics, Agent technology

摘要： 分布在因特网上的物流资源具有地理分散和职权自治的特性，资源结构和接口难以统一。该文以网格、Agent和增量挖掘技术为基础，提出了不通过资源整合就能够实现全局信息挖掘的方法。分析了基于Web的物流资源网格系统，将其划分为物流域的集合实施分级管理。提出了新的面向网格的信息挖掘模型并设计了域内动态资源挖掘算法和域间请求式资源挖掘算法。该模型解决了不同物流系统之间的信息挖掘难题，算法中引入的增量挖掘技术提高了域间资源挖掘效率。

关键词: 资源网格, 增量挖掘, 物流, Agent技术

CLC Number: 

• TP391