Research Of Web Mining Application In E-Government

With the process of design and implement of E-Government web site and DongFeng Yulon office automation system, it would be appropriate to meet users' needs if web personalization could be implemented according to users' interests. Web mining was one of the most effective methods for web personalization. Web usage mining of users' browsing behaviors could implement personalization recommendation effectively.There was a thorough research on the difference of web content mining, web structure mining and web usage mining. It demonstrated that web usage mining was significant for the personalization of E-Government web site. According to the general process of web mining, the personalization plan was put forward to be the same with E-Government web site. The plan analyzed advantages and shortcomings of collecting web data at server, proxy and client side for obtaining users' browsing behaviors and a preprocessing method to log data for web usage mining was introduced, including data cleaning, user identification and session identification. It reduced the pages which could not reflect users' interests and enhanced the precision of the data through the analysis of the pages structure and the time between two pages. The characteristics of the rule for recommendation were analyzed and a plan of one support number association rule was introduced in application of web site. Moreover, an association rule mining algorithm based on the idea and structure was designed. Meantime, another idea of the weighted page support was suggested. The method using above made recommendation more accurately. It was designed that personalization recommendation service of an E-Government web site by using this plan.