Background: Decision-making trial and evaluation laboratory (DEMATEL) is a practical and concise method to deal with the complicated socioeconomic system problems. However, there are two defects in original DEMATEL. On the one hand, the traditional expert preference expressions cannot reflect the hesitation and flexibility of expert, on the other hand, the experts’ weight is usually expressed as the equivalent weight which cannot reflect the scientific weight on behalf of the experts’ academic background, capability experience, risk preference and so on. To solve the above problems, a novel Group DEMATEL decision method based on hesitant fuzzy linguistic term sets (HFLTSs) is proposed.
Method: Firstly, this paper presents that experts make their judgement on the causal relationship of factors by using a linguistic expression closed to human expression, which can be easily transformed into HFLTSs. Next, the hybrid weight of experts are calculated on the base of the initial HFLTSs direct influence matrix(HDIM) according to the hesitant degree and distance between two HDIMs. And the aggregation of each expert’s information is introduced by possibility degree. Then, the new group DEMATEL decision method based on HFLTSs are constructed. Finally, an illustrative example is given and analyzed to demonstrate the effectiveness and validation of the proposed approach. Results: This paper demonstrate the heterogeneity of decision experts and the hesitation degree of expert information representation must be taken into account when determining the interaction of factors in complex systems by DEMATEL method. Conclusion: This paper constructs the new amended group DEMATEL which provides a new way to deal with the integration of each expert’s information by the hybrid weight and possibility degree. The methods provide references for determining the importance of complex system factors more scientifically and objectively.Keywords: Decision-making trial and evaluation laboratory, group DEMATEL, hesitant fuzzy linguistic term sets, expertsweight.