User modeling for pedestrian navigation services

Abstract

Human-centered and user-adaptive systems are at the heart of the Design for All and Ambient Intelligence concepts. Evidently, user models are necessary "ingredients" of such systems. We present a user model for navigation systems (mainly pedestrian), which is based on relevant human wayfinding and navigation theories. We represent this model through a Semantic Web ontology and show how it can be incorporated in an indoor navigation system called MNISIKLIS, which enables personalized path selection. Moreover, we propose a method for learning rules for the classification of users based on Inductive Logic Programming. © 2009 Springer-Verlag Berlin Heidelberg.