A two-stage similarity clustering-based large group decision-making method with incomplete probabilistic linguistic evaluation information

In recent years, probabilistic linguistic term set (PLT) is widely used in large group decision making (LGDM) for its integrity. However, the complexity of probabilistic linguistic LGDM and the large span of experts’ profession cause two problems. On the one hand, it is difficult for all experts to give complete evaluation information in the form of PLTs. For this, we propose an expertise-based probabilistic linguistic evaluation information complement method. First, we identify authoritative experts under each attribute through professional hesitation and professional consistency. Then, we establish an optimization function to obtain the optimal missing value through the expectation score of authoritative experts and the linguistic term using habit of pending experts. On the other hand, the similarity between two experts cannot be fully represented by the sum of the distance of expert evaluation value. For this, we propose a two-stage similarity measurement method and introduce the distance weighting process, which not only measures the similarity between two expert evaluation values, but also measures the difference in degree of distance between two experts under different attributes. Finally, we apply this LGDM method to hot dry rock exploration site selection in southeast coast of China.