<![CDATA[This study is motivated by the significant potential of Jakarta’s traditional culinary sector in tourism, which is not yet supported by interactive digital identification media for tourists. This research aims to implement a Convolutional Neural Network (CNN) algorithm to automatically classify traditional snacks as part of a smart tourism system. A primary dataset consisting of 2,600 images across 14 snack classes was collected from Ciracas Market and Setu Babakan in April 2025. Data preprocessing was conducted using augmentation techniques, followed by model development using transfer learning with two architectures, ResNet50 and EfficientNetB0. The experimental results show that ResNet50 achieved the highest accuracy of 99.08% with a loss value of 0.0239, outperforming EfficientNetB0, which decreased to 94.48% accuracy. The best model was deployed in a Streamlit-based web application that provides interactive information, including the name, history, and ingredients of traditional snacks. This system facilitates tourists in recognizing Jakarta’s local culinary heritage and supports the implementation of smart tourism.]]>
Copyrights © 2026
