Probabilistic linguistic term sets (PLTSs) can express the qualitative information of decision makers more accurately in the complicated linguistic setting. However, the existing comparison methods for PLTSs cannot compare some special PLTSs. To tackle this problem, a novel score function based on the concentration degree of an PLTS, called ScoreC-PLTS, is proposed. Additionally, the existing distance measures may distort the original information and lead to unreasonable results. Therefore, a novel probability splitting algorithm is proposed to preprocess PLTSs, based on which, a novel generalized hybrid distance is proposed for PLTSs. Moreover, a novel multiplicative analytic hierarchy process (MAHP) based on ScoreC-PLTS is proposed to determine the weight vector of attributes. Based on the generalized hybrid distance and MAHP, two novel TOPSIS-ScoreC-PLTS and VIKOR-ScoreC-PLTS methods are put forward to handle multi-attribute decision-making problems with PLTSs. Afterwards, an illustrative example concerning the selection of children English educational organization is solved using the proposed TOPSIS-ScoreC-PLTS and VIKOR-ScoreC-PLTS methods. In this example, four indicators are developed and the superiority of our studies is verified by comparing with the previous TOPSIS and VIKOR methods.

概率语言术语集（PLTS）可以在复杂的语言环境中更准确地表达决策者的定性信息。但是，现有的PLTS比较方法无法比较某些特殊PLTS。为了解决这个问题，提出了一种基于PLTS集中度的新颖得分函数，称为ScoreC-PLTS。此外，现有的距离度量可能会使原始信息失真，并导致不合理的结果。因此，提出了一种新颖的概率分裂算法对PLTS进行预处理，并在此基础上提出了一种新颖的广义混合距离。此外，提出了一种基于ScoreC-PLTS的新型乘性层次分析法（MAHP）来确定属性的权重向量。基于广义混合距离和MAHP，提出了两种新颖的TOPSIS-ScoreC-PLTS和VIKOR-ScoreC-PLTS方法来处理PLTS的多属性决策问题。之后，使用建议的TOPSIS-ScoreC-PLTS和VIKOR-ScoreC-PLTS方法解决了有关儿童英语教育组织选择的说明性示例。在此示例中，开发了四个指标，并且通过与以前的TOPSIS和VIKOR方法进行比较来验证我们研究的优越性。