Study Of Adaptive Web Site Based On Artificial Neural Network

In order to solve the problems of network information flooding and lack of individual network services and promote the high development of network services, it becomes a current hot issue that whether a web site could automatically adapt different user's requirements or not. Web mining is the sub field of data mining. Features of the variety and dynamitic of data source breaks through the restriction of traditional data mining, and with the development of web, web mining gets more and more attention and application. Adaptive web site is one of the application branches of web mining. An adaptive web site can automatically adjust its structure and representation so to meet different users visit requirements through using web data mining technique to find the useful information of users. Thus research on adaptive web site can resolve the requirement and need contradiction between visitors and web service.An executive framework of adaptive web site based on theory of artificial neural network setting up model is put forward, which applies Extensible Markup Language (abbreviation for XML) technology. The feature of this model is that it aims at performance of individual services, and gets user's favoritism by using algorithm of difference transmitting based on artificial neural network, meantime using the web structure mining and user access pattern to perform web mining. Individual configuration of web users is implemented by XML technology during site adapting phases. Compared with other similar systems, these system emphases on user's individual service and it emerges web structure and user access pattern.This paper introduces the design and implementation of adaptive web site in detail, including analysis of user favoritism on basis of difference transmitting algorithm and etc. It also gives a demo of applying part of framework on a practical web site and gets fairly good experimental result.