No abstract

In this paper, an approach for reorganizing Web sites based on user access patterns is proposed. Our goal is to build adaptive Web sites by evolving site structure to facilitate user access. The approach consists of three steps: preprocessing, page classification, and site reorganization. In preprocessing, pages on a Web site are processed to create an internal representation of the site. Page access information of its users is extracted from the Web server log. In page classification, the Web pages on the site are classified into two categories, index pages and content pages, based on the page access information. After the pages are classified, in site reorganization, the Web site is examined to find better ways to organize and arrange the pages on the site. An algorithm for reorganizing Web sites has been developed. Our experiments on a large real data set show that the approach is efficient and practical for adaptive Web sites.

In order to run a successful website, it is significantly crucial for website owners to understand users' intentions and desires. By capturing these information, they can provide better service and enhance marketing strategy to achieve this goal. Web usage mining (WUM) is an application that can help people to explore the useful patterns of users' browsing usages. Traditionally, it discovers knowledge from Web log data. However in some websites, they offer a service that users can select or enter some required criteria from fields, and these information will be saved online. These criteria show the intentions or desires of a certain object required for this user. Interested persons can enter queries or browse categories to find these posted cases. In this paper, clustering method is applied to group similar users based on these collected required criteria in a website. When dataset is huge, it is difficult to find the characteristics of individual group. Thus association rule mining is applied to each cluster. The generated rules can be inferred to identify the interests and characteristics of users in each group. Finally, marketing decision can be made especially for each group's users.

No abstract

No abstract