Batik is one of Indonesia's cultural heritages with complex and diverse patterns, possessing high artistic value and deep philosophy. Manual classification of batik patterns requires time and depends on expert knowledge, making the process inefficient. This study aims to develop a batik pattern classification model using Convolutional Neural Network (CNN) with a data-driven approach, enabling automatic and accurate pattern recognition. The dataset used consists of 4,284 batik images divided into five pattern classes: Kawung, Lereng, Ceplok, Parang, and Nitik. In this research, the CNN model was developed by using transfer learning techniques with MobileNetV3 pre-trained on a large dataset. The training process involved data augmentation to enhance the model's robustness against variations in batik patterns. The evaluation was conducted by measuring the model's accuracy and loss. The results show that the CNN model achieved an average accuracy of 93.42% on the training data and 93.88% on the testing data. This research demonstrates that the data-driven approach using CNN is effective for batik pattern classification, providing more accurate results compared to manual methods and offering an efficient solution for the digitalization of the batik industry. The developed model can serve as a foundation for broader applications in cultural preservation and the advancement of artificial intelligence-based technology.