Using Abduction in Markov Logic Networks for Root Cause Analysis

In this paper we propose an approach for calculating the most probable root cause for an observed failure in an IT infrastructure. Our approach is based on Markov Logic Networks. While Markov Logic supports a special type of deductive inference, known as maximum a posteriori inference, the computation of the most probable cause requires abductive reasoning. Abduction aims to find an explanation for a given observation in the light of some background knowledge. In failure diagnosis, the explanation corresponds to the root cause, the observation corresponds to the failure of a component or service, and the background knowledge corresponds to the dependency graph of the infrastructure extended by potential risks. We apply the method for abduction proposed by Kate et al. to extend a Markov Logic Network in order to conduct abductive reasoning [1]. We illustrate that our approach is a well suited method for root cause analysis by applying it to a sample scenario.