Problem Definition

When foreigners consider visiting Korea, they typically focus on major cities like Seoul, Busan, or Jeju. However, Gwangju, with its stunning tourist attractions such as mountains, parks, and museums, remains an underrated gem. As a group of foreign students with limited knowledge of Gwangju, we often face challenges in planning trips that align with our interests, budget, and available time. To address these challenges, we propose developing a website that assists not only foreign students in Gwangju but also those in other cities, as well as international visitors, in discovering the hidden beauty of Gwangju. By leveraging machine learning models, including Gradient Boosting Regressor and K-means clustering, the platform will predict places users will enjoy and recommend nearby restaurants and hotels for more conveniences based on their preferences, budget, and time constraints. In addition to model selection, we will focus on feature engineering to improve the accuracy and relevance of recommendations. To further refine and personalize recommendations, the platform will incorporate continuous learning mechanisms. This approach will allow the system to update and improve its models in real-time based on user interactions and feedback. This system aims to enhance the travel experience by providing personalized and comprehensive trip recommendations, making Gwangju more accessible and enjoyable for all visitors.

System Design

Our system aims to suggest nearby and preferred locations to users, helping them maximize their enjoyment without wasting leisure time and money. This system leverages React.js for the frontend, Flask for the backend, and MySQL for the database, with code managed through GitHub and deployed via AWS amplify/AWS EC2. This architecture ensures a robust, scalable, and efficient solution to provide personalized trip recommendations.

Figure 1: Overall System Design