在税收征管系统中，纳税人走逃失踪往往导致欠税税款无法追回，为提高税务机关在欠税管理方面的效率，可利用多维关联规则挖掘对欠税行为的多个维度进行分析。本文利用R语言改进Apriori算法，且在关联规则评价上，使用客观评价与主观评价相结合的方式，同时利用衍生规则的方法删除冗余规则，得出纳税人走逃失踪行为与其他维度的有效关联规则信息，为税务机关提供有效的决策信息，提高欠税管理的风险防范水平。%In tax collection and administration system, the overdue tax by disappearing taxpayers always can-not be paid for. To improve the efficiency of overdue tax management, the paper applies multidimensional associa-tion rule mining to analyze multiple dimensions of overdue tax behavior. It uses the improved Apriori algorithm in R language. In the evaluation of association rules, in addition to using the combinative way of objective evaluation and subjective evaluation, it uses the method of derivative rules to delete redundant rules, and obtains the effective association rules between the behavior of the disappearing taxpayers and their other dimensions. It provides effec-tive decision-making information for tax authorities, and improves the capability of risk prevention in overdue tax management.
展开▼
