Causal network topology analysis: Characterizing causal context for risk management.

Abstract

The ways that risk assessments are commonly performed in organizations have limitations that undermine their quality. They typically focus on individual risk events one at a time but are weak at integrating their relevant causal context, into decision-making processes. Network topology analysis has previously been applied to address this weakness through quantitatively characterizing the importance of the causal interactions of risk events. However, there remains a lack of both clarity and consistency in terminology, methods, and interpretation of the results of this approach. This paper presents and formalizes causal network topology analysis, a methodology that contributes to (1) characterizing the causal context of a risk event to inform its management, (2) articulating the ontological concepts underpinning a repeatable topology network analysis, and (3) justifying the selection and usage of network metrics for this purpose. The theory and methodology are discussed, and an exemplar application to a mining project feasibility study is presented.