Goal-aware data management for retrieval and recommendations

Human activity is almost always intentional. By understanding why user-generated events are happening and what purposes they serve, a system can offer a significantly improved and more engaging experience. Analyzing user actions such as clicks can reveal patterns and behaviors. However, understanding the goals behind these actions is a more challenging issue since goals cannot be easily captured. In this thesis, we present a unified, multi-disciplinary viewpoint for goal management that covers many different cases where goals can be used and techniques with which they can be exploited. We are dealing with the problem of selecting and recommending items that will fulfill, or contribute towards the fulfillment of user goals. Specifically, we suggest a goal-aware solution for the problem of finding related forum posts to a post at-hand within a user-community. Our solution is based on the detection of intention shifts captured through text features. Furthermore, we suggest a general purpose solution for recommending items of any kind by leveraging knowledge about goals that users managed to fulfill in the past through interactions with these items. The items to be recommended are selected considering whether and how they will contribute towards the fulfillment of a number of goals.