Expert System for Troubleshooting Computer Network and Performance

Student: Sani Gambo Ibrahim (Project, 2025)
Department of Computer Science
Umaru Musa Yaradua University, Katsina, Katsina State

Abstract

This work “hybrid rule-based machine learning expert system for local area networks (LAN) troubleshooting” was done to provide analysis into modern approach to the development of expert systems, and to apply this approach to the computer local area networks (LAN) problem domain. This was based on the need to provide support for system and network administrators to quickly resolve minor network problems in the absence of a network solution specialist or expert. The decision tree machine learning model combined with a rule-based engine as a hybrid model was explored and shown to be effective in building decision support systems whose possible outcomes are not ambiguous.  To build a system that is also responsive, the machine learning model caters for cases in which rules are not defined, while the rules engine handles cases in which rules are pre-defined. This approach provides more comprehensive troubleshooting for common computer local area network issues encountered on daily basis in a small network environment, and as such can be assumed as a good fit for building a more robust modern expert system for general network troubleshooting.            

Keywords
network expert system machine learning local troubleshooting approach model rules
Filters
Institutions