Xml-based And Svm Web Text Mining Research

With the development of Internet, information of Internet increase quickly, one of the instances we face now is the user's aspiration of obtaining needful information quickly and exactly, the other one is huge amount of information and complexity of information structure, these things make difficult process information. To solve the conflict, Web mining techniques provide an approach, research of Web mining is developing now, it need to research about theory and technique. The dissertation mainly researches about Web text mining techniques.The dissertation researches the Web text mining in detail according to the process of Web text mining, constructs a Web text mining model based on eXtensible Markup Language (XML) and Support vector machine (SVM). the Web text mining model based on XML and SVM possesses function of Web text preprocessing and Web text mining, its advantages are reducing amount of data step by step by fixing on authority pages, XML technique, feature selection in order to obtain term gather that can express text correctly and reducing dimension of high- dimension data by support vectors machine, refines data that text mining need to process.The dissertation focuses on research of process and technique of Web text preprocessing, the dissertation indicates structuring the information in Web pages by XML, and then express these texts by format that computer can deal with, extract useful information for text mining, reduce the amount of data, form a text feature database for text mining.