Abstract
In this paper, we identify ambient social interactions as an important direction of research in Ambient Intelligence. To enable intelligence for ambient social interactions, we build knowledge-based systems to aggregate information from heterogeneous sources, including physical environments, users' social tasks and available services. Such knowledge-based systems enable intelligence not only for individuals, but also for user groups. We propose a general framework for aggregating information from different sources. A novel representation, Entity Notation (EN), is suggested as a starting point for the purpose of connecting all information to knowledge-based systems, which, in its turn, offers good possibilities to support Semantic Web technologies for development of smart functions for ambient social interactions. To demonstrate the usability of our framework and representation, we present an implementation of a map reminder service, and report analysis based of it.