A dynamic user profiling technique in a AmI environment

Currently there are many services that can assist the human being in his decision making. Many of these services provide aid as consistent as possible attending the characteristics and preferences of a user, which compiled results in a person's profile. Without profiles, systems could not provide a coherent aid to a user. In addition we can consider the profiles as the basis of recommendation systems. Paramount with cognitive helping systems, that provide decisions and recommendations these actions can more accurate and user driven. However, the profiles need to be updated over time, as a human being changes of preferences or beliefs, the profiles also need to adapt to dynamic environments. It is introduced a project that applies the Bayesian Networks and Case-Based Reasoning techniques to create and modulate user profiles in a coherent and dynamic way, using stochastic models and high-level event relations and characteristics to devise an accurate suggestion of activities the user can perform, being integrated in an Ambient Assisted Living Project.