Enterprise Contextual Intelligence

Information overload is an increasing challenge for the enterprise knowledge worker. Traditional information retrieval, i.e. Search-based approaches for knowledge management in the enterprise are under strain because users do not have the time to search, often they are not even aware that material relevant to they current needs exists. Neither do they have the time to track the various external news feeds that are increasingly becoming available, however personalised they might be. We submit that relevant content should be pushed to users based on detecting their current needs from the context of their current activities as sensed via their behavioural footprint, such as their posts, emails, as well as search queries. In this paper we describe a generic ´enterprise contextual intelligence´ (ECI)framework based on an ontology-driven probabilistic graphical model for push-based context-aware knowledge recommendation in an enterprise. We illustrate our ECI framework with a motivating practical business example and differentiate it from other context-aware search and recommendation approaches.