A two-stage personalized feedback mechanism considering dynamic interactive behavior under social network in large-group emergency task scheduling schemes selection

Emergencies such as natural disasters frequently occur around the world. Large-group emergency evaluation (LGEE) based on Earth observation satellite data is a new model that can potentially reduce the negative impacts of such emergencies. However, in real situations, the different interest preferences and diverse educational backgrounds of evaluators in the evaluation group may cause conflictive preferences, leading to low confidence in the results, which will affect the effect of LGEE. Therefore, this paper aims to devise a two-stage personalized feedback mechanism driven by the dynamic interactive behavior under a social trust relationship to coordinate evaluators' opinions with conflicting opinions in the LGEE. Firstly, we introduce a family of probabilistic linguistic trust-propagation operators to obtain the complete trust relationship among the group using Archimedean t-norms and then determine the weights of evaluators and the weights associated with each pair of evaluators. Secondly, a Louvain method is used to divide the entire group into several subgroups to reduce computational complexity. Next, a two-stage personalized feedback mechanism is built to manage the consensus levels of intra and inter subgroups by describing the dynamic interactive behavior during the evaluation process. The first stage is to assist evaluators in achieving the consensus within the subgroup. The second stage is devoted to reaching an agreement of each subgroup to the global group. Finally, a case study of the selection problem in emergency task scheduling schemes is undertaken to verify the practicality and validity of the proposed method, followed by some comparative analysis and discussion.