PART 1. Predict the category of a structure
The trained image classification model was evaluated through random image prediction and visual inspection, confirming that it can accurately recognize different categories of historical structures. By learning distinctive architectural features, the model demonstrates strong potential for real-world heritage analysis beyond standard accuracy metrics.
From a broader perspective, this AI-driven approach has significant implications for heritage preservation and tourism enhancement. Automated classification of historical structures can assist government agencies and cultural organizations in monitoring monument conditions, identifying structures that require maintenance, and prioritizing conservation efforts. Additionally, such models can be integrated into digital tourism platforms to provide intelligent recommendations, improve visitor engagement, and deliver educational content tailored to specific heritage sites. Overall, the use of AI contributes to sustainable heritage management while enriching tourist experiences through data-driven insights.
PART 2. Perform exploratory data analysis and develop a recommendation engine
A recommendation system was developed using tourist rating data to provide personalized place recommendations. By computing similarity between users based on their rating behavior, the system identifies tourists with similar preferences and recommends attractions that the target user has not yet visited. This approach effectively captures shared interests without relying on explicit user profiles.
The recommendation model enhances the tourism experience by helping travelers discover relevant destinations aligned with their tastes. From a broader perspective, such personalized recommendation systems support smarter tourism planning, encourage exploration of diverse attractions, and contribute to sustainable tourism growth by distributing visitor traffic more evenly across locations.