The problem that is often faced by investors in selling/buying stocks is the difficulty in analyzing a dataset of stock prices in large quantities. This analysis aims to predict the rise or fall of stock prices based on data obtained. To assist investors in determining buying/selling decisions on stock analysis based on technical and equipped with classification techniques in data mining. This study analyzes the performance of the J48 Decision Tree algorithm in the Waikato Environmental Software for Knowledge Analysis (WEKA) version 3.8. 2 for PT. Harum Energi Tbk.(HRUM). The results showed in the testing data, the percentage of testing on data without normalization was higher by 87.3 (non-aggressive) and 88.8 (aggressive) compared to normalized data 84.2 (non-aggressive) and 85 (aggressive). The biggest stock profit generated is in non-aggressive type data without normalized by 48.75 or Rp. 48,750.00.