Bayesian Network Based Software Diagnosis Expert System
Shunkun Yang · 2012
Through the demonstrated external phenomenon or some input data, most software failures can be analyzed to help the in-depth understanding and backtracking the root source of the fault. Base on bayesian network, we designed a fault analysis expert system, which can realize the collection of fault case, strcture learning and parameter learning of the diagnosis network. This diagnosis expert system can analysis the possible fault cause and fault module, and output the dynamic diagnosis strategy given the inputs of fault phenomenon. And finally it can be achieved the multi-perspective shallow reasoning, deep-level reasoning and dynamic fault reasoning.