Text Data Mining For Applied Research In Information Monitoring

With the development of net technology, WEB information is speeding up incredibly. It is true that most of the information are useful except for the harmful minorities. For the net information monitor sections, It is the most important and hard work to stop the harmful ones in time.In these days, the information monitor sections use information search means. It is based on the way of key-word searching. It is often to get too much worthlessness and it isn't easy to describe the aim by key-word fabrication. That results too much work and most of all, the monitor becomes very inefficiently.According to this, the writer improves the information searching means. Data mining technology is used in WEB information monitor system to improve the searching ability to clean up the net environment.In the article, the writer expounds the basic idea of the monitor system design and the application of data mining .in the view of the basic idea: firstly, collecting the text information from WEB for building the information base. Secondly, Building text categorization model, and training the model by a great many harmful information samples data. Thirdly, collecting the text information from the WEB , weeding out the hamful ones by the well-trained text categorization model. Lastly, training the model by the new text in often with the time of using to make the categorization result good more and more.Also Data Mining of text audio categorization model and WEB information collecting are stated completely in the article. To web information collecting, WEB audio categorization model ,text structure mining are studied, such as Chinese departing words , vector space model and text audio categorizating model (supporting vector method and KNN method) and the like.The use of text data applying in information monitor system will be a great useful help in the study and application life.