User Profiling and Recommendation System

A personalization system can be intended as a computer-based application which monitors the behavior of the user, to create and manage his/her profile, with the purpose to select only relevant products/items, within a very large repository, or to support other applications and services in adapting to the specific needs of the user.

Such a system provides recommendations that are typically based on user profiles, which are built and refined by taking into account user actions. More challenging, and not yet fully explored, is the domain of context-aware recommendation. Context information can be acquired by processing data from sensors available on the mobile device or through explicit user input.   

Based on this principle, we have developed UTravel, an effective user profiling and recommendation system for smart mobility. The suitability of recommended points of interest (POIs) is inferred from user-provided evaluations, according to the principles of collaborative filtering.