A Situational Resource Rating System

Recommendation technologies are considered a major technological trend in both industrial and academic environments. This growing interest was highlighted by, e.g., the Netflix prize which generated an intense competition. Recommender systems are crucial to support users and help them by suggesting resources relevant at a given instant. On the other hand, these systems are a core piece of e-commerce web sites, since they aim at generating more sales by encouraging users to buy more items. However, recommender systems are often designed to work with very specific types of resources, and they hardly take into account the current user's situation. In this paper, we present our approach to augment an existing recommender system with a situation model. On top of this model, we define a situational interest measure to estimate a user's interest for a resource, which we demonstrate with a prototypical implementation.