Abstract: With the rapid development of the economy and society and the continuous deepening of reform and opening up, many lawsuits have increased significantly. Some effective legal documents have not been enforced, which is called “difficult to execute”. For a long time, “difficult to execute” has not only been a pain point that plagued the judicial enforcement of the people's courts but also a hot spot that has been strongly reflected by the people and has attracted widespread attention from all walks of life. In judicial execution, it is difficult for the judge to retrieve and extract high-value information timely and accurately due to a large number of historical property concealment cases, few judges, and complex property concealment clues. In addition, under the influence of such factors as case handling experience, life experience, knowledge accumulation, information asymmetry, and personal bias, judges have systematic reasoning biases in judicial decision-making, which to some extent reduces the accuracy of judgment. If the traditional off-line execution mode is still adopted for one-by-one screening and judgment, it will not only lead to low execution efficiency but also be difficult to ensure the trial quality of each case. When multiple execution cases are received, the judge needs to further determine the priority of the execution cases, and accordingly allocate judicial resources reasonably.
　To address the “difficult to execute” problem of the court, this paper applies the fuzzy clustering analysis method to the judicial execution field. The data type of clustering is extended to hesitant fuzzy linguistic information, and a complete evaluation system of the hidden property behavior of the person subjected to execution is constructed. By estimating the quantified probability of hidden behavior of the person subjected to execution and ranking the execution cases, it provides decision support and judgment basis for the judge to determine the focus of investigation and control. At the same time, it avoids the disadvantages of experiential judicature and improves the objectivity and reasonability of judicature decisions. This paper analyzes the limitations of the existing hesitant fuzzy linguistic distance measures, defines the hesitance degree of hesitant fuzzy linguistic information, and develops a new method for distance calculation of hesitant fuzzy linguistic information. Using the idea of maximizing deviation to determine the optimal attribute weight, a hesitant fuzzy linguistic agglomerative hierarchical clustering algorithm based on maximizing deviation is proposed. The effectiveness of the evaluation system and the clustering algorithm is verified by the cluster analysis process of the executor in the context of hesitant fuzzy linguistic decision information.
Solving the difficult execution problem is a complex systematic project. In future work, it is necessary to refine the research questions and application scenarios, such that execution cases will be divided into enterprise execution cases and people's livelihood execution cases, and decision-making situations will be divided into individual decision-making and group decision-making. The applicability and pertinence of the model will be improved.

Key words: hidden property, executor, hesitant fuzzy linguistic term sets, maximizing deviation, distance measure, clustering algorithm

摘要： 针对法院“执行难”现状,将模糊聚类分析方法应用于司法执行领域,聚类的数据类型拓展为犹豫模糊语言信息,构建了一套完整的被执行人隐匿财产行为评估体系。通过估测被执行人隐匿财产的量化概率,为执行法官确定查控重点提供了决策支持。对当前的犹豫模糊语言距离测度存在的局限性进行了分析,给出了犹豫模糊语言犹豫度的定义和新的犹豫模糊语言距离计算方法。采用离差最大化思想确定最优属性权重,提出了一种基于离差最大化的犹豫模糊语言凝聚式层次聚类算法。犹豫模糊语言决策信息环境下的被执行人聚类分析算例验证了评估体系和聚类算法的有效性。

关键词: 隐匿财产, 被执行人, 犹豫模糊语言术语集, 离差最大化, 距离测度, 聚类算法