Automated Summarization of Educational Videos

Abstract

With the growing availability of educational videos online, students often face the challenge of efficiently extracting key information without spending excessive time taking notes manually. This thesis presents a web-based application that leverages artificial intelligence to automatically convert lecture videos into structured, high-quality notes, making learning faster and more effective. The system allows users to submit a video link. If subtitles are available, the application retrieves them; otherwise, it extracts the audio and generates a transcript using AI speech-to-text tools. The transcribed text is then summarized and organized into clear, concise notes using AI models like GPT, with sections, key points, and optional timestamps. Users can manage their notes through editing, translating, and downloading in various formats, enabling a personalized learning experience. Built with a React frontend and a FastAPI backend, the application demonstrates how modern web technologies and AI can work together to enhance educational outcomes. The results show that students can save time while capturing the most important concepts from video lectures, improving study efficiency and accessibility.