The Study And Application Of Fuzzy Data Mining Technology Based Internet

Presently, the Internet has been the most important resource of theinformation. It is real useful to help users quickly and correctly seek the informationof their needs, which has been the hot point in recent years. I research it, one is to usethe technology of data mine to improve the correction rate of index and categorization.The other is really to analyse some algorithm through studying these problems and tryto improve the algorithm. The main research work is as follow: One is to study the theory of auto-classification. Text auto-classification includesthe auto-collection categories and auto-assignment categories, here only study theauto-assignment categories. Text auto-assignment categories include trainingprocedure and classifying procedure. It is the most important to choose the trainingalgorithm and documents in the training procedure. We adopt fuzzy algorithm. Only ifbuilding the base of better training documents, we get better categorizationachievement. So we find enough number training text from lots of big specializedportal site. The producing model in the training procedure is the base of thecategorization procedure. Our system has two levels, the first level use the charactercategorization and the second one use the word categorization. The first improves thespeed of categorization and reduce the correct rate and the second is against. The other is to explore how to improve the classification algorithm. Based thepredecessor's research, I choose fuzzy algorithm and deeply study it, then I adjust itslightly and make it adapt the classification require. In the future, I hope to continue the research and deeply study it in the term ofinformation categorization.