A Cascade Mining Algorithm Based on Chinese Keywords Web Mining

摘要

Security content filtering of World Wide Web is one of the important tasks among network security. The lower precision of Web mining based on keywords is a common fault, especially when those grouchy persons used active disturbing methods to cheat and bypass various filters. To filter these few but purposively or malicious Web pages the first thing is the classifier design. Therefore, a cascade mining algorithm was proposed, which consisted of one cascade classifier operator and three mining components, including jamming mining component, Bopomofo mining component and complicated characters mining component. After mined with the three components, the temporary mining result of a Web page would be dealt with the cascade classifier operator and got the last resu the classified Web page was normal or malicious. Experiments and analyses demonstrated that the cascade mining algorithm can solve the poser of how to classify for Chinese Web page simply, effective and precision.