Personalizing relevance on the Semantic Web through trusted recommendations from a social network

Personalization efforts to date have centred on presenting web users with novel items by predicting what they may find relevant. This approach has utility where the user is unsure of exactly what they are looking for, but not where they have a particular information need to satisfy or a particular item to locate. Furthermore, by operating purely on a predefined database of users and items, systems using this approach represent closed worlds and offer poor scalability to new data sets. To address these limitations we propose a technique for personalizing relevance in information seeking activities, based on an understanding of how people seek information and recommendations from their social network. We then describe technical work in progress, based on Semantic Web technologies, that aims to realize this perspective.