User Profile Mining and Personalization of Web Services

Web Services are defined as software systems designed to support interoperable machine-to-machine interaction over a network using standardized XML messages. Since user’s expectations and requirements constantly change, it is important to include their preferences in offering of web services. A user profile, used in web service personalization, is a structured construct containing information both directly and indirectly pertaining to a user's preferences, behavior and context. Effective personalization requires services to build and maintain accurate models of a customer’s preferences, interests and background through a user profile. Building effective user profiles can benefit from different research contributions in different areas, including security, statistical prediction and mining etc. In this paper we focus on the dynamic evaluation of user profiles for personalization of web services based on service usage log.