A Unified CRITIC-Similarity Hybrid Weighting and Aggregation Model for Multi-Criteria Decision Making

Abstract

Decision-making involving multiple competing criteria often relies heavily on subjective expert judgement, which may reduce the transparency, consistency, and reliability of the final decision. Many existing approaches lack an objective mechanism for determining criterion weights and provide limited flexibility in accommodating variations in expert preference structures. For instance, although Kamis et al. (2018) proposed a structured Group Decision-Making (GDM) framework, it does not explicitly incorporate objective weighting techniques or examine the influence of different similarity measures and aggregation operators on ranking stability. To address these limitations, this study proposes a hybrid Multi-Criteria Decision-Making (MCDM) framework that integrates the Criteria Importance Through Intercriteria Correlation (CRITIC) method for objective criteria weighting. The CRITIC method determines criteria weights based on contrast intensity and intercriteria correlation, thereby reducing subjective bias and reflecting the inherent characteristics of the decision data. Furthermore, three similarity measures, namely Cosine Similarity, Manhattan Similarity, and Additive Reciprocal Preference Similarity (ARPS), are incorporated into the GDM model to capture different aspects of expert preference similarity. The resulting similarity scores are aggregated using the Weighted Arithmetic Mean (WAM), Induced Ordered Weighted Averaging (IOWA), and Weighted Ordered Weighted Averaging (WOWA) operators. The proposed framework is demonstrated through a case study on hybrid energy system selection considering technological, economic, and environmental criteria. The results show consistent and stable alternative rankings across different similarity measures and aggregation operators, indicating improved robustness and reduced sensitivity to subjective judgement. Overall, the proposed framework provides a comprehensive and reliable decision-support model by integrating objective weighting, similarity-based preference modelling, and flexible aggregation techniques for complex multi-criteria decision-making problems.