Discovering chronicles with numerical time constraints from alarm logs for monitoring dynamic systems

  • Authors:
  • Christophe Dousson;Thang Vu Duong

  • Affiliations:
  • France Telecom CNET, Lannion Cedex, France;France Telecom CNET, Lannion Cedex, France

  • Venue:
  • IJCAI'99 Proceedings of the 16th international joint conference on Artifical intelligence - Volume 1
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

We address the problem of knowledge acquisition for alarm correlation in a complex dynamic system like a telecommunications network. To reduce the amount of information coming from telecommunications equipment, one needs to preprocess the alarm stream and we propose here a way to acquire some knowledge to do that. The key idea is that only the frequent alarm sets are relevant for reducing the information stream: we aggregate frequent relevant information and suppress frequent noisy information. We propose algorithms for analysing alarm logs: first stage is to discover frequently occurring temporally-constrained alarm sets (called chronicles) and second stage is to filter them according to their interdependency level. We also show experimental results with an actual telecommunications ATM network.