"Classification of Export Transaction Risks Based on Incoterms Attributes, Payment Terms, and Logistics Capacity: A CRISP-DM Framework"

Abstract

This study aims to classify the risk level of export transactions into two categories: High Reliable (reliable) and Low Risk, based on Incoterms attributes, payment terms, and logistics capacity. The study is motivated by the need for relevant authorities to identify reliable export partners in support of national export development policies. The research adopts the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework, which consists of six phases: business understanding, data understanding, data preparation, modeling, evaluation, and deployment. The dataset comprises Indonesian exporter data integrated with Harmonized System (HS) code data. The Random Forest algorithm was employed to develop the classification model. The results demonstrate that the proposed model performs well in distinguishing reliable exporters from those with higher transaction risks. Among the evaluated variables, monthly production capacity and the value of goods per kilogram were identified as the most influential factors in determining exporter reliability.