A RAG-Based Framework for Smart Travel Itinerary Generation
摘要
Global connectivity and tailor-made experiences are rapidly transforming the travel industry. This research involves the application of Retrieval-Augmented Generation (RAG) to improve travel recommendations. Our system takes the effort out of planning travel, by bringing together generative AI, real-time data integration, and user-focused de sign. It features a React.js front-end and Python Flask back-end, integrating Google Gemini, OpenWeatherMap APIs, and web scraping for real-time updates on destinations, weather, and lodging. Travel plans dynamically adapt to unexpected weather, roadblocks, or hotel changes. Some of its features include secure authentication, real-time weather updates, personalized recommendations based on user preferences, budget, and travel duration, among others. Experimental confirmation for the scalability and adaptability in providing current travel features.