Chronicle Learning and Agent oriented techniques for network management and supervision

This paper presents some aspects of the “Reseau fute (Smart Net)” project, whose aim is to introduce Artificial Intelligence techniques (such as machine learning and multi-agent systems) in network management and supervision, in order to help the processing of the large volume of events notifications received by network management operators. We provide experimental and theoretical results on learning patterns called chronicles in order to design a machine assistant to network operators. Theoretical results investigate different levels of help that could be brought by the operator to the assistant. The tests were performed in two distinct realworld situations. They showed the circumstances under which chronicle learning is possible without the help of the operator or another assistant.