A method of tri-layer network modeling under strategic objectives for project portfolio risk assessment

Risk assessment for project portfolio (PP) is pivotal in aligning strategic objectives (SOs) with organizational value propositions. Despite its significance, the impact of project portfolio risks (PPRs) on achieving SOs remains understudied, potentially causing inaccurate assessment. Applying a Bayesian network (BN) to model a tri-layer (PPR-PP-SO) network, this study bridges the gap and forges a direct connection between PPR assessments and SOs. Initially, the PPR criteria are determined, followed by constructing a BN structure to clarify relationships within the three-layer network. To mitigate data scarcity limitations in BN, Spherical Fuzzy Set and Dempster-Shafer theory are introduced to acquire tri-layer BN parameters. Utilizing propagation analysis of the tri-layer BN, the expected impact of risks on SOs is measured to reveal PPR criticality. A numerical example corroborates the model’s functionality and feasibility. The results indicate the criticality of one PPR varies substantially under different SOs, implying the importance of pinpointing critical risks specific to each SO.