Abstract: In the decision-making process, linguistic terms can express the preferences of decision makers.The probabilistic linguistic term sets (PLTS) can be not only given directly by experts, but also given by the evaluation information of different groups' decision-makers in a large group.so PLTS has a very important role in the decision-making process. However,there are few methods for PLTS multi-attribute group decision making in the exiting researches. The aggregation process and consistent measurement are important factors in group decision making.At present, the research of PLTS has just started. The related research on PLTS does not pay attention to the problem of consistency measurement. The existing aggregation methods are less applicable to PLTS multi-attribute group decision making,and the aggregation methods are not only computatively more complex, but also the aggregation results are unreasonable. The multi-attribute group decision-making method of PLTS needs further study. Therefore, the purpose of this paper is to propose a new consistency measure and aggregation method based on the existing research, and to establish a PLTS multi-attribute group decision method which considers the consensus measure method. Its main feature is that decision makers' weights are unknown,we consider the consistency of decision maker to make decisions. Firstly, decision makers' weights are determined by defining a consensus measure of PLTS, and proposed a aggregation to aggregate the evaluation information of each decision maker. Finally, ranking PLTS by the degree of possibility formula. In addition, in order to clarify the effectiveness and advantages of the method in this paper, the comparison between the method in this paper and the existing method shows that the proposed method avoids a large number of operations and the final ordering which contains probability information can illustrate the decision result more clearly. What's more,the proposed method could beused for reference to the aggregation of probabilistic language evaluation information, and can provide an effective tool for multi-attribute group decision making problems with unknown weights of decision-makers.

Key words: multi-attribute decision-making, probabilistic linguistic term set, aggregation method, degree of similarity, consensus measurement