A granulation of linguistic information in AHP method using modified particle swarm optimization algorithm

Multi-criteria decision making methods have been of the most popular and practical areas in recent years. Among these methods, the AHP method has been considered by many researchers due to its unique features and has been used in numerous studies. This method facilitates decision-making for the decision-maker by allowing pairwise comparison between alternatives using language expressions. But this feature causes inconsistency in the decision matrix. One of the sources of inconsistency is the use of a predetermined scale to convert linguistic variables into quantitative values due to the different intellectual background of experts and their information about the problem. Therefore, the purpose of this study is to provide a customized scale for converting linguistic variables using the granularization of linguistic variables, which is in line with expert opinions and reduces the inconsistency. The unique feature of this framework is that the distribution of cut-off points is not known in advance and is determined according to expert opinions. To optimize the proposed model, a particle swarm optimization metaheuristic algorithm is used, which is modified to adapt to the specific characteristics of the problem. The results show the good performance of the proposed framework in reducing incompatibility.