Context-aware collaborative filtering system: Predicting the user's preference in the ubiquitous computing environment

In this paper we present a context-aware collaborative filtering system that predicts a user's preference in different context situations based on past experiences. We extend collaborative filtering techniques so that what other like-minded users have done in similar context can be used to predict a user's preference towards an activity in the current context. Such a system can help predict the user's behavior in different situations without the user actively defining it. For example, it could recommend activities customized for Bob for the given weather, location, and traveling companion(s), based on what other people like Bob have done in similar context. © Springer-Verlag Berlin Heidelberg 2005.