Post-harvest Soybean Meal Loss in Transportation: A Data Mining Case Study

A poultry company in Indonesia has a problem, i.e., losing raw mate-rial, the so-called Soybean Meal (SBM), during transportation from the port to the factory. To reduce material loss, the company created a raw material transport (RMT) system, which recorded the time and activities during loading-unloading and transporting the material from the port to the factory warehouses. Therefore, this study aims to mine the data on the loss of raw materials through RMT. The application used is Orange data mining to find the relationship between lost material and other attributes, create clusters, and classify the standardized lost. The clustering exhibits two classes, namely, the standard and non-standard condi-tions. The classification process uses five different algorithms. The random forest algorithm was chosen because it produces the second-best AUC value and can produce a classification visualization through a decision tree. This classification process also produces rules based on the decision tree.