The Application Of Data Mining In Filtering Harmful Information From Internet

With the rapid development of Internet in China in recent years, it has played a more andmore important role in promoting the development of both our economic and society, improvingthe whole people's scientific and cultural level, and enhancing the socialistic spiritual civilization.But at the same time, internet security management now faces with a lot of new problems. Someenemies abroad and some domestic criminals commit all kinds of severe crimes through Internet,which harms the state security, social stability, and the construction of socialistic spiritualcivilization, meanwhile blocks the healthy growth of adolescents. These crimes widely arouse thedispleasure of all classes in society.According to the " The 16th Statistical Report for Internet Growth" published by CNNIC,there had been 103 million internet subscribers and 677,500 websites up to June 30th, 2005. Indealing with so many websites, the traditional way of content inspection via manual work cannotdetect specific internet information in given time, even in huge cost of manpower and capital. Inthis condition, it is necessary to develop a type of computer software to help us searching andanalyzing information automatically, thereby providing an efficient technical tool for detection.In this article, I mainly do further researches on Web mining, Text ming, the mechanism ofsearch engine, WebBrowser components and Mshtml components, furthermore design "a systemfor searching and collecting harmful messages in the Internet". The main issues in designing andapplying this system are as follows:1.After analysing the technology of data mining, web mining and text mining, I put forward away of feature extraction and keyword search in Interactive columns such as BBS and chat rooms.2.After analysing the structure of websites and the constitution of BBS, and setting websitesearching strategy and using DFS, we can achieve searching websites and BBS quickly.3. .By analyzing architecture of IE and the fundmental function of WebBrower componentand Mshtml componet, analyzing the page structure of chat rooms by using HTML element.Afterall these we make it reality that we can automatically login chatroom and also can watch thechatroom information dynamically.4.I also talk about the mechanism and model of searches of internet search engine. Byanalyzing the character of outcomes from search engines such as Google and Baidu, we can getthe data's universal code of these search engines.