基于数据挖掘的海关风险分类预测模型研究

Customs Risk Classification and Forecasting Model Based on Data Mining

海关业务每日产生的海量记录中蕴藏着数据"金矿"有待进一步挖掘,为加强海关风险识别的准确性,让大数据的价值进一步得到显现,本文采用数据挖掘分类分析的方法,对历史报关单数据进行分析,根据其查获情况,将有查获与否作为分类标号,建立分类模型对历史报关单进行分类,提取相关的规则,揭示数据中隐藏的规律并运用其规律进行预测,为报关单的风险评估预测提供参考。

The daily record of mass production of customs business contains the data “ gold mine” to be further excavated. In order to strengthen the accuracy of customs risk identification and make full use of the value of big data,classification analysis of data mining is adopted to analyze the data of histor-ieal customs declaration. The records are tagged as hit or not hit according to its hit result. The classifi-cation model classifies the historical declarations, extracts the relevant rules, reveals the hidden rules in the data and uses the rules to predict. The result could be applied in the risk assessment and forecast of the declarations.