Abstract:

To answer the advocation of “Going global” strategy by China’s central government， Chinese manufacturing companies have been increasingly and deeply participating global competition in various international markets. However， continuously emerging environmental dynamic changes and risks， such as those from unfair competition， regional politics， pandemics and etc.， have been requiring to build robust supply networks for adaptation. The practical requirements thus also have been highlighting the crucial role of comprehensively evaluating resilience capability in supplier selection， which accentuates to simultaneously and comprehensively consider economic， environmental and social indicators as well as resilience ones.According to extant literatures， the focal problems of comprehensive evaluation of suppliers’ resilience capability exhibit multi-faceted complex characteristics， such as holding semi-structured or low-structured problem definitions， involving participation of multiple stakeholders in decision-making processes， incurring inaccurate assessments and decision hesitancy among decision makers. To this end， with acknowledgement of the multi-faceted complex characteristics above-mentioned， it focuses on investigating effective group decision making approaches for comprehensive evaluation of suppliers’ resilience capability by contributing to introducing in this paper： （1） an effective hybrid expression tool for eliciting decision-makers’ assessments more objectively and comprehensively； （2） fuzzy information measures for avoiding potential information distortion in comparison to widely-accepted traditional ones； （3） rational methods for objectively obtaining unknown weighting vectors； （4） processual modelling methodologies for group decision making to accommodate typical situations of suppliers’ resilience capability evaluation.To solve these problems listed above， firstly， a novel expression tool called as dual hesitant unbalanced linguistic term set （DHULTS） has been defined. Secondly， to avoid potential information distortion that is usually brought about by traditional hesitant element extension methods， a least common multiple （LCM） based extension mechanism has been introduced to develop correlation coefficient-oriented measure for DHULTS， by use of which a weighting method for obtaining unknown experts’ weights has further been established. Thirdly， grounded on the above methods， corresponding group decision making approaches have been constructed with regard to two specific situations in practices： （i） target supplier（s） for benchmarking could be derived in advance； （ii） decision making units do have different opinions on prioritization among evaluative attributes. Finally， case application and numerical studies have then been conducted to verify effectiveness and superiority of our approaches.

Key words: dual hesitant unbalanced linguistic term set, correlation coefficient, weighting model, supplier’s resilience capability evaluation