---
title: "DANTES: An Expert System for Real-Time Network Troubleshooting"
authors:
  - "Robert Mathonet"
  - "Herwig Van Cotthem"
  - "Leon Vanryckeghem"
doi: ""
url: "https://ijcai.org/proceedings/1987/105"
pdf: "https://ijcai.org/proceedings/1987/0105.pdf"
venue: "Proceedings of the Tenth International Joint Conference on Artificial Intelligence"
year: 1987
pages: "527-529"
track: "Knowledge Representation"
keywords:
---

# DANTES: An Expert System for Real-Time Network Troubleshooting

Robert Mathonet, Herwig Van Cotthem, Leon Vanryckeghem

*Proceedings of the Tenth International Joint Conference on Artificial Intelligence*, pages 527-529.

## Abstract

Today's ~omputu networks ar• IDrg• and ~omplu. Their dayto-day operation and malntenan~• ~an benefit from the support of an •xpert system, mainly as an aid in troubleshooting. Network troubleshooting has ~hara~tuisti~s, like in~omplete data, high rate of events, simultan•ous presence of several problems, whi~h raise inter•sting probl•ms in the d•velopment of an expert system. DANTES is an expert system designed to provide r•al-time assistance to network operators. This paper presents the system and stresses the dndopment issu•s that are peculiar to network troubleshoot i ng . Of parti~ular importance are : performance of inferen~e in r•al-time, multi-problem handling, and ~ons ideration of time in reasoning and revision of belief. Dealing with su~h issues and especially with real-time effi~iency is primarily a question of system design. This has impli~ations for th• knowledg• base organization, reasoning me~hanism, and re~ording of dedu~tions