Evaluation of a rule-based approach for context-aware services

In previous work we have proposed a rule-based platform to support context-aware service development. Rules are used to detect relevant situations in the users’ context and to trigger actions in response to context changes. Rule-based solutions offer flexibility with respect to service maintainability, since rules can be modified and added at runtime. However, rule-based systems are often criticized for being monolithic systems that only perform adequately for rather static datasets, which would imply that these systems are not capable of providing the performance and scalability required by large scale context-aware services. In this paper we show that this criticism does not always hold by evaluating our distributed rule-based implementation and demonstrating the suitability of this approach for highly dynamic context-aware services