Selection and ranking of activities in the social web

The paper presents a framework for the selection and ranking of activities proposed to users in a social networking service, as for example Facebook’s “social events”. Our proposal takes into account the peculiarities of this application area, considering a variety of factors including the spatiotemporal relations between activities and between activities and users, and the social opinions weighted by the user’s interest in the specific themes and type of an activity. To this aim, we describe a semantic model for the involved entities, and especially for the available activities. We describe a process that can be easily configured according to user preferences, extended to take into account additional ranking factors, and adapted to include existing recommendation/ranking strategies. Finally, we specifically introduce ranking factors that can be used within the specific model and process we describe. The results of a preliminary evaluation show the effectiveness of our approach.