Research On Web Text Mining

In the past few decades,the development of WWW made it become the largest public datasource in the world.It contains all kinds of messages.WWW contains a few billon interconnected web documents.We are confronted with the problem that we can get very little from a great deal of net message.In usual,we mainly pay attention to the message of Web Text.The message of web is complicated,distributed and refreshed.It also contains many unusal message and the web user is complicated.These characters web and the problem that it is very hard for us to get web message make us need to mine Web Text for getting more message. we must resolve the problem that how to search for more correct and effective message.The Rising of web mining technology bring us with hopeness of solving the problem.Web Mining is being the period of development.We disscussed the theory of web text mining and deeply research the critical technology of web text mining.The main contect is as follows.(1)We discussed the aim and means of web text mining and define the basic concept.We also introduce web mining and the place it can be put in use.We design the chart of web text mining system.(2)We research the pretreatment technoloty ot web text and analyse the procedures of web processing.We designed the chart of text extracting.We research the algorithm of word cutting.We introduce a few methods of using feature words to substitute for text.We also research the methods of decreasing dimension.We analyse all the procesure of text processing.(3)We introdue the procesure of text mining.We research the important technology of web text classfication.Web research the technology of web text clustering.At last,we introduce how to estimate the mining results.(4)We propose a kind of clssfication method of attribute weight.We also propose the method of the better SVM web text classfication.We analyse the mining result that proves the method is better than previous method.