Integrating user data and collaborative filtering in a web recommendation system

Abstract

Web-based applications with a large variety of users suffer from the inability to satisfy heterogeneous needs. Systems should be capable of adapting their behavior to the user’s characteristics, such as goals, tasks, interests, which are stored in user profiles. Filtering techniques are used to analyse profile data and provide recommendation to the users to help them navigating in the site and retrieving information of interest. We describe here the approach we have adopted in FAIRWIS (Trade FAIR Web-based Information Services), a system that offers on-line innovative services to support the management of real trade fairs as well as Web-based virtual fairs. The approach is based on the integration of data the system collects about users, both explicitly and implicitly, and a classical collaborative filtering technique in order to provide appropriate recommendations to the user in any circumstances during the visit of the on-line fair catalogue.