Answers are grounded in this paper and cite their sources.
Traditional green supplier evaluation methods frequently rely heavily on expert judgment and subjective assessments, which constrain their objectivity and comprehensiveness. The current study developed a hybrid data-driven framework that integrates Random Forest (RF) with Multicriteria Decision-Making (MCDM) methods to assess suppliers across economic, environmental, and social sustainability dimensions. The framework incorporated RF for identifying influential sustainability criteria with DEMATEL, ANP, and MOORA-AS for modelling interrelationships, weighting, and ranking suppliers. Results revealed that Technology (C2) and Supplier Collaboration (C8) were the most critical factors influencing sustainability outcomes. Meanwhile, DEMATEL identified Warehouse Management (C1) and Water Usage Efficiency (C3) as crucial drivers affecting other criteria, where ANP assigns the highest weights to C2 (23.1%) and C8 (22.3%). The MOORA-AS ranking disclosed Supplier 2 as the best performer, which served as a benchmark for sustainable practices. Moreover, improving criteria such as C3 (Water Usage Efficiency) and C4 (Sourcing) could indirectly enhance social performance by influencing C8, which demonstrated strong interdependencies among sustainability dimensions. The proposed model enhances analytical objectivity and reduces reliance on subjective weighting in traditional evaluation methods, although expert judgement remains involved in criteria definition and qualitative indicator coding. This framework enables companies to develop more resilient, transparent, and sustainable supply chains by linking data-driven analysis with sustainability principles.
No takes yet. Share an insight, caveat, or question.
Runtuk et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: