CARMS: Clustering based approach for recommendation in mediation systems

摘要

Nowadays, where the number of information is increasing exponentially, there is necessity to filter and deliver relevant information to the user. For this purpose, information systems are more and more interested to provide users with personalized content and services by searching through the large volume of available information, the most correspondent to user's needs. This approach is called the personalization of information system and is based on the concept of user profile. The personalization can show up either by adapting the interface, filtering information or proposing recommended content to the user. In this work, we propose a general user profile similarity function toward a clustering based approach to personalize mediation systems by recommendations.