Hesitant fuzzy linguistic preference consistency-driven consensus model with large-scale group interaction measure for venture capital investment selection

Recently, consensus-based large-scale group decision making (LSGDM) has been widely interactive with the study of social network, clustering and trust-based concepts. This study develops a novel hesitant fuzzy linguistic preference consistency-driven consensus model with interaction measure for large-scale group decision makers (DMs) in social networks. Firstly, directed social network is constructed by measuring the similarity between incomplete hesitant fuzzy linguistic preference relation (HFLPR) matrices. Community detection method is further conducted to categorize DMs into several communities. Secondly, driven by exploring the consistency of HFLPR matrices and interactive trusts between DMs, a novel optimization model is established to estimate the missing elements. Thirdly, the 2-order additive fuzzy measures of different coalitions between divided communities for capturing their fully or partially interactions are derived by a consistency-based optimization model. Accordingly, the attitudinal Choquet integral operator is employed to aggregate preferences into the collective one. Fourthly, a consensus improving mechanism is devised to achieve the unanimous agreement of DMs characterized by the bounded confidence. Personalized and specific adjustment scales obtained by investigating interval consistency of HFLPRs are provided in support of DMs’ modifications. Finally, an illustrative case on syndicated venture capital investment selection is conducted and related simulation analyses are performed to elucidate the feasibility and validity of the proposed methods. The comparisons with other approaches reveal the superiority and improvement of our proposal.