Study Of Web Mining Methods

Data Mining is the result of people keeping studying and exploiting database technology. Data Mining is that people extract information or mode, which is connotative, unknown and valuable, in large database or data warehouse. That is a new field of being applied value in database studies, combines multi-fields' theory and technology, such as database, artificial intelligence, machine learning, statistics and so on. Web Mining is one of hotspot of Data Mining technology in artificial intelligence fields, which implements some function as web access mode, web structure and rule, dynamic search for web content, is a more defiant subject. This paper discusses web content (text) mining mainly.Firstly, this paper summarizes data mining and web mining technology, analyzes and studies the feature of web data, compares XML with traditional database and selects XML document to save available data. Secondly, it presents the metod of implementing this subject based on the task of web mining, which combines neural network with boosting arithmetic to text classification. Compared with single neural network method, it can improve identification rate of samples and classified accuracy much more.Now, the system has experimentally run and the result is excellent. It has reached its goal of experiment and study, which establish bases to the further research of web mining.