Personalized individual semantics based consensus reaching process for large-scale group decision making with probabilistic linguistic preference relations and application to COVID-19 surveillance

•A new definition of distance between PLTSs is proposed.•DM’s difference in linguistic terms is reflected by personalized individual semantic.•The CRP (consensus reaching process) considers DMs’ willingness.•Construct minimum preference adjustment models to obtain the adjusted PLPRs.•The CL (consensus level) of the adjusted DM is increasing in each adjustment round. This paper develops a new personalized individual semantic (PIS) based consensus reaching process (CRP) for large-scale group decision making (LSGDM) with probabilistic linguistic preference relations (PLPRs) and applies to the selection of COVID-19 surveillance plans. Firstly, considering that a linguistic term means different things to different decision makers (DMs), a new definition of distance between probabilistic linguistic term sets (PLTSs) is defined considering the PIS of a DM. Then, a consistency-driven optimization model is built to determine the PIS of linguistic terms in a PLPR of a DM by maximizing the consistency of a PLPR. DMs’ weights can be acquired through a programming model by minimizing the distance between the individual semantic and the collective semantic. Besides, a clustering algorithm based on PIS is devised for dividing DMs into several subgroups with similar semantic. Based on the obtained subgroups, the corresponding moderators can be identified by DMs with unacceptable consensus level (CL). Several minimum preference adjustment models are constructed to obtain the adjusted PLPRs. These models not only can sufficiently consider the willingness of DMs to modify their preferences but also can improve their CLs in CRP. Finally, an illustration example of selection of college COVID-19 surveillance plans is offered to demonstrate the proposed method and to verify the effectiveness of the proposed model.