摘要： 介绍了一个剪报资料处理和检索系统EPCutter(Electronic Press Cutter)， 它包括3个子系统：剪报资料处理子系统、网上智能代理子系统和全文数据库检索子系统。网上智能代理子系统利用数据挖掘技术和智能自主代理技术，自动从Web上挖掘出用户感兴趣的信息；全文数据库采用了字索引和词索引相结合的索引检索方法，从而大大提高了检索速度、查全率、查准率。此外还提出了一个统计模型，可对剪报来源作出评价以辅助用户决策。

关键词: 电子剪报, 数据挖掘, 智能代理, 全文数据库

Abstract: This paper presents a clipping information processing and retrieval system EPCutter (electronic press cutter), which includes three subsystems: resource processing module, intelligent Web agent module and fulltext database retrieval module. The intelligent Web agent module utilizes data mining and auto-intelligent agent technology in order to obtain some useful information automatically on Web. In this fulltext database, index by single word and index by phrase combine to form the hybrid approach which possesses high performances in rate, accuracy and allsidedness for information retrieval. In addition, a statistic module is given to evaluate information sources of every description and reach a decision about source selection.

中图分类号: 

• TP399