Risk Identification-Based Association Rule Mining for Supply Chain Big Data

摘要

Since most supply chain processes include operational risks, the effectiveness of a corporation's success depends mainly on identifying, analyzing and managing them. Currently, supply chain risk management (SCRM) is an active research field for enhancing a corporation's efficiency. Although several techniques have been proposed, they still face a big challenge as they analyze only internal risk events from big data collected from the logistics of supply chain systems. In this paper, we analyze features that can identify risk labels in a supply chain. We propose defining risk events based on the association rule mining (ARM) technique that can categorize those in a supply chain based on a company's historical data. The empirical results we obtained using data collected from an Aluminum company showed that this technique can efficiently generate and predict the optimal features of each risk label with a higher than 96.5% accuracy.