Prediksi Status Pesanan Menggunakan Metode Classification C.45 Pada Toko Stuftech.Id di Shopee

Abstract

The increase in transactions on e-commerce platforms such as Shopee necessitates the development of accurate order status prediction systems to optimize services and reduce order cancellations and delays. This study aims to build a classification model for predicting order status (completed or canceled) at StufTech.Id store using the C4.5 algorithm. The dataset consists of transaction attributes such as payment method, shipping location category, and shipping cost. The classification process was conducted using RapidMiner through preprocessing, decision tree generation, and model evaluation stages. The analysis shows that the “Shipping Region Category” attribute has the highest information gain and was selected as the root node. The resulting model achieved an accuracy of 86%, with 100% recall for completed orders but only 6.67% for canceled ones. These findings indicate that the C4.5 algorithm is effective in predicting successful transactions but requires improvement in identifying potential cancellations. Implementing this model can help business owners make proactive operational decisions.