Collaborative recommendation of ambient media services

Abstract

Ambient intelligence environments are technologically augmented surroundings that aim to provide personalized services to the users based on their context. Identifying these services for the users has become an increasingly challenging task. The overwhelming number of services in the ambient environment has made the selection and management of services even more challenging. To address this problem, researchers have proposed several techniques, such as creating a user model and selecting services based on that model; applying rule-based approach to match the relevant services; utilizing a combination of user's profile, context, interaction history and service reputation to select the best services for the user, and so on. Most of these techniques obtain the preference of a user based on his/her own interaction and profile and do not consider the power of collaborative selection approach. In this paper, we propose to use the collaborative recommendation technique to select services for a user based on multiple users' interactions and profile. Accordingly, we demonstrate the potential of the proposed approach through preliminary experiment.