Bayesian networks in project management: A scoping review

Abstract

To support project management (PM) activities, many methods and techniques have been proposed. However, little information exists on how these methods and techniques are used. This is particularly the case with the technique of Bayesian networks (BNs). The purpose of this scoping review is to categorize and to describe how BNs are applied in PM. To this, 102 journal articles published between 2004 and 2021 were analyzed. First, we describe the nature and evolution over time of the articles. Second, based on a specifically developed framework, we classified the articles by project type (Construction & Infrastructure; Software & IT; Engineering & Manufacturing; and Others), project aspect (Challenges & Risks; Context & Process; and Outcomes), reasons for using BNs (Description; Prediction; and Prescription) and types of BNs (Basic BNs; Combined BNs; and Extended BNs). In general, the findings highlight continued evolution of the literature on the subject, mainly in the last five years (2017–2021). Furthermore, thanks to the classification framework, we have suggested to researchers avenues for future research. Practitioners could also use it as a basic tool to implement a PM program.