Based On Web Log Data Mining Research

Develop continuously along with the Internet, the data scoops out technical research and application to also become a little bit hot problem more and more, is focused attention fully in international local academic circles and information industry field.How apply a data excavation's technique in the Web, the exhumation is useful from the daily record of Web server, important knowledge(include mode, rule...etc.), become an importance research and application realm of data excavation and knowledge detection, this is according to the data of the Web daily record to scoop out.Article the logarithms made to elaborate generally according to the excavation first, such as the definition, function of data excavation, classification etc..Immediately after the logarithms carried on the research of system according to an important daily record excavation of the branch-Web of excavation.The process that the author's circumambience scoops out to the Web is this clues, particularly is to prepare processing and mode to discover these two stages launched detailed thorough of study.Aim at a few steps that the data prepares a processing stage, if the data cleans, the customer identify, the conversation identify, path's complement's etc. carried on the study of point, and give it concretely carries out calculate way.End, aim at one of the applied realms of Web excavation, author to from orientation of the Web station ordered to launch to discuss more and in detail, and put forward ordering from the orientation station of structure strategy etc. series problem.Put forward a kind of connection matrix of the data structure-customer/page(URL_UserID) which is applicable to Web daily record to scoop out, mean customer's interview to page information, and make use of to gather a type, page to gather type and multifarious interview path according to the customer of the connection's matrix etc. calculate way, carry out customer and page to gather a type and customer thus of multifarious interview path.