面向交友网站的Web日志关联规则挖掘及系统实现-计算机软件与理论专业毕业论文.docxVIP

ABSTRACT With the development of online dating site, there are obvious commercial profits in making advertisement purposefully for the dating users. For the purpose of improving the site popularity, reducing the operating costs and gathering more users, it is essential for dating site to use Web log mining technology for the access records of the website, to find out the users browsing patterns, understand the users interests and preferences of visiting. Finally，by improving the service of the site and the channel and pattern of advertising, let more users know the site and trust it. Aiming at the actual demand from the Web site of Ocean of women, the algorithm of association rules is applied to Web log mining, while the specific process and key technologies were studied. The Web log mining system is implemented, which meets the requirements of the site. The main contents include: The existing Web log mining techniques are studied, analyzed and compared; A new way to get the site topology automatically is proposed. In the general pretreatment process of the data of Web log mining, the topology of the site is usually used to identify users and session and supplement path. Now, usually people create the topology structure of the site artificially according to the structural features of the site or download structural information of the site from the Web server by offline browser software, then process it offline manually for constructing the topology structure. Moreover, when we confront the dynamic site, offline browser software can not get complete structure information of the Website. For this reason, a new method to get the site topology automatically by software is put forward. A mining system for Web log based on association rules is designed and implemented for the online dating site. The system includes 4 functions: getting site structure automatically, data pretreatment, pattern discovery, pattern analysis. At first, preprocess the raw log data, then combin