Application And Practice Of Development Based On Collaboration

World Wide Web is an open global distributed network. Resources on the net do not haveuniform structures and can not be managed easily. So it is difficult to find some informationfrom the Internet. Web page classification can avoid the disorder of web information greatly.It can help user to locate the needed information and sort information. At the same time, thedevelopment of the Internet requires higher quality information search services, and it doesnot meet users' need only based on the content of web page. So it is necessary to developChinese web page classification tools that are fit for our country, and it can assist users tomanage and control web information better. This dissertation studies web page classification deeply aimed at above conditions. Itsummarizes text classification and web page classification based on the content, and comes upwith a new concept that is character classification. At the same time, this dissertation analysesthe character classification such as the sense, the feasibility, the concrete algorithms and so on.Finally, it introduces a kind of application to research engine. The contributions of thisdissertation are as follows: 1. It summarizes the procedure of text classification and the structural character of webpage, and elaborates the algorithms of web page based on the content including KNN, SVM,Bayes, decision-making tree and so on. 2. A new concept that is character classification of web page is presented. It analyses thefeasibility and necessity of this technology after studying a lot of structural character of webpage and comes up with the concrete algorithms of character classification includinghypertext, hyperlink, file format and so on. 3. This dissertation compares content classification and character classification, andpoints out their sameness and differences. For example, both are same at sense, managingobject, algorithms idea, development field, and are different in implication, detail procedure,development status and so on. 4. The application of optimizing research outcome is realized in this dissertation by twokinds of result classification agents of different frame. One is an agent based on inquiryoptimization, and the other is an agent based on result optimization. They are compared topoint out the proper scope.