10-06-2017, 06:51 AM
A knowledge discovery framework for the construction of Community Web Directories, a concept that we introduced in our recent work, applying the personalization to web directories. In this context, the Web directory is viewed as a thematic hierarchy and personalization is done by constructing user community models based on usage data. In contrast to most of the work on web mining, usage data discussed here pertains to user browsing throughout the Web, rather than a particular Web site, resulting in a high degree Of thematic diversity. To model user communities, we introduce a novel methodology that combines the navigation behavior of users with thematic information of web directories. Following this methodology, we improve the clustering and the probabilistic approaches presented in previous works and also present a new algorithm that combines these two approaches. The resulting community models take the form of Community Web Directories. The proposed personalization methodology is evaluated both in a specialized directory and in a general purpose Web directory, indicating its potential value to the Web user. Experiments also evaluate the effectiveness of different mechanical learning techniques in the task.