Trust Risk Test-Based Group Consensus with Probabilistic Linguistic Preference Relations under Social Networks

Xinru Han,Jianming Zhan,Zeshui Xu,Luis Martinez

DOI: https://doi.org/10.1109/tfuzz.2024.3377126

IF: 12.253

2024-01-01

IEEE Transactions on Fuzzy Systems

Abstract:Social networks are gaining importance as an important application landscape for machine learning techniques. Currently, it is prevalent thinking to introduce social networks into group decision-making research, despite exhibiting promising performance, there remain significant challenges, including 1) trust risk is often overlooked, and 2) group opinions are excessively relied upon. In this regard, this article introduces the trust risk test for group consensus (SN-TRT-GC) by using probabilistic linguistic preference relations under probabilistic linguistic term sets. The proposed method ensures that the expert's competence matches the discourse as much as possible by virtue of the trust risk value. Moreover, opinion evolution relies on the dual attributes of similarity and trust, which promote the exchange of opinions among experts and greatly accelerate the consensus reaching process. Empirical investigations of satisfaction surveys demonstrate the validity and superiority of the presented SN-TRT-GC technique.