Usage of Artificial Intelligence in Smart Tourism: A CNN-based Landmark Classification for Diyarbakir Province

The integration of artificial intelligence (AI) in the tourism sector has emerged as a transformative approach to enhance the travel experiences and promote regional attractions. This study explores the application of AI in smart tourism by developing a convolutional neural network (CNN)-based landmark classification system for Diyarbakır Province, a culturally rich region in South-East Turkey. The method specifically focuses on classifying images of five popular landmarks in Diyarbakır, leveraging state-of-the-art AI techniques to enhance regional visibility and tourist engagement. To achieve this, a pre-trained CNN model, namely AlexNet, was fine-tuned for the landmark classification task. By adapting the parameters of AlexNet to the specific dataset, the model was optimized for improved feature extraction and classification accuracy for totally 5 classes (places). The proposed framework was evaluated using a carefully curated dataset, yielding a remarkable 97.5% classification accuracy on the test set. The performance of the proposed model highlights the reliability and effectiveness of the methodology in accurately identifying landmarks, even with complex architectural and environmental features. These results can concretely contribute to both academic research and practical applications by providing a lightweight and accurate model that can be embedded into mobile travel applications or digital tourism platforms for real-time landmark recognition, enhancing tourist engagement and regional visibility.