Two-stage group decision making methodology with hesitant fuzzy preference relations under social network: Multiplicative consistency determination and personalized feedback

Abstract

Some social network group decision making (SNGDM) researches may overlook two issues: (1) the impact of consistency on decision reliability and decision makers’ (DMs’) status in social network, and (2) DMs’ personalization during consensus reaching process. In response to these two issues, a two-stage SNGDM model is proposed for hesitant fuzzy preference relations (HFPRs). In the first stage, multiplicative consistency determination and improvement of HFPRs are achieved. For complete and incomplete HFPRs, several programming models are constructed to classify consistency types. When it comes to inconsistent HFPRs, an inconsistency fixing method is designed to ensure the original HFPRs as much as possible. With the improvement of DMs’ consistency, their status in social network will be dynamically enhanced. In the second stage, two programming models are devised to elicit the reasonable individual priority weight vector in consistent HFPR. Further, minimum adjustment cost-based personalized feedback mechanism is invented to attain the optimal personalized adjustment coefficient and recommendation, and achieve the harmonious unity of individual personalization and group consensus. Finally, the superiority of our proposal is verified by comparative analysis and illustrative example.