ABSTRACT
This chapter addresses personalization of multimedia content in the Future Internet from a recommendation perspective. It presents research results towards providing context-aware recommendations in the Future Internet, based on the semantic modeling of knowledge and recommendation approaches, mixing both semantic and fuzzy processing for better personalization. The chapter focuses on a Hybrid Ad-Hoc Network environment, constituting a common communication infrastructure today, where people share and manipulate multimedia content on both fixed and mobile network nodes. A context-awareness platform based on the Content-Centric Network paradigm is presented, together with ontologies for context and situation, thanks to which contextual elements can be formalized. The resulting context-awareness platform feeds a knowledge-based recommender with contextual data used to weight and filter user interests. The chapter examines the remaining drawbacks of the most adopted recommendation approach, namely collaborative filtering. It highlights the need for semantic, fuzzy preferences and context-awareness toward the building of more powerful recommender systems, able to make suggestions closer to users’ expectations.
