More and more services rely on user localization, yet presenting a location information in an intuitive form hasn't really been studied to our knowledge. We present a method to calculate the best way to express a user's location, given a priori modeling of their knowledge about people and places, so that users can get "at a glance" information about the location of their relatives. Our approach is a first attempt at designing such a method. We present the results obtained, and discuss the many ways to improve it and to generalize the method to other context dimensions.
Download Full PDF Version (Non-Commercial Use)