Adaptive Augmented Reality (A2R) is an emerging technology that can support users in their daily life with useful information for their activities which is really adapted to the user's characteristics, to the environment where the activity is taking place, and to the current context. However, one of the problems identified is the lack of a formal definition of the models required and the logical architecture for the development of A2R systems. As a first step to this goal, our aim is to propose a detailed definition of the content of the User Model required for this type of systems. We explored state of the art ontologies for user modelling, and propose a set of significant user characteristics to be modelled. Also, we present an initial architectural model for this type of systems.
