基于Web信息挖掘的商业分析系统设计  

Design of commercial analysis system based on web information mining

作　　者：沈洁[1] 林颖[1] 赵敏涯[1] 陈志敏[1] 

机构地区：[1]扬州大学计算机科学与工程系,江苏扬州225009

出　　处：《计算机工程与设计》2006年第1期62-65,共4页Computer Engineering and Design

基　　金：江苏省高校自然科学基金项目(02KJB520013)

摘　　要：随着Web上信息的快速增长,如何将潜藏于非结构化文档中的商业信息有效提取并分析服务于商业管理已成为新的研究热点。利用现有的Web信息挖掘技术,针对原始数据的混合异构性,提出信息块多主题分割的方法,在建立的商业领域实体名字典指导下对商业信息进行抽取和分类,并引入一种信用评级机制,构造了一个基于Web信息挖掘的商业信息分析系统(CABWIM),实验结果表明系统能有效地将散落在Web中潜在的商业信息抽取并加工整理,形成真正有实用价值的商业信息。With the rapid growth of information on the web, how to efficiently extract and analyze latent commercial information from semi-structured pages to improve the commercial management has become a hot research issue. Using the existing technologies of web information mining, a new method that could segment topics from information block in order to settle the problem of complex data structure was proposed. With the direction of business entity name dictionary, this method extracted and classified business information. And also a mechanism of credit estimating to insure precision was introduced. Finally, a commercial information analysis system based on web information mining （CABWIM） was constructed. Experiments show this system can efficiently extract and tidy latent information to valuable commercial information.

关 键 词：WEB信息挖掘 包装器 DOM树 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]