15th European Conference on Artificial Intelligence
|July 21-26 2002 Lyon France|
This paper is dedicated to the chronicle recognition approach used to design an evolution monitoring system for supervising dynamic systems for which time information is relevant. We propose to extend and also to unify the chronicle representation through event counters. The main motivation of such an extension of the chronicle representation raises from alarm processing: counting the occurrences of alarms can be useful to evaluate the severity of a problem and also to discriminate some kind of faults. The paper describes the representation and the corresponding algorithms for processing this extension according to the purpose of supervision in terms of performance.
Keywords: temporal reasoning, knowledge representation, diagnosis
Citation: Christophe Dousson: Extending and Unifying Chronicle Representation with Event Counters. In F. van Harmelen (ed.): ECAI2002, Proceedings of the 15th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2002, pp.257-261.