Abstract:
The rapid growth of digital learning platforms has created a demand for intelligent systems that can deliver personalized, engaging, and efficient educational content. Lection Ai is an AI-driven educational tool designed to automatically generate mini-lectures, quizzes, and learning materials tailored to individual learners’ knowledge levels. The system leverages Large Language Models (LLMs) to transform user-uploaded educational content or topic inputs into structured lecture slides, summaries, and assessment materials with minimal human intervention.
Lection Ai follows a modular architecture with a Flutter-based mobile frontend and a Python Flask backend. The backend integrates an LLM (via Ollama) to generate lecture content and quizzes in a structured JSON format. A machine learning–based evaluation mechanism analyzes quiz performance, considering both accuracy and completion time, to classify learners into beginner, intermediate, or advanced levels. Based on this classification, the system dynamically adapts the depth, complexity, and presentation style of the generated lecture content.
A key research contribution of Lection Ai is the enhancement of lecture materials through the automated integration of relevant visual content. The system retrieves educational images using a curated dataset and the Google Custom Search API when required, embedding them into generated PDFs and PowerPoint presentations. This approach aims to improve learner engagement and knowledge retention. Additionally, the system focuses on optimizing processing time and scalability, enabling efficient handling of large content inputs and concurrent users.
Overall, Lection Ai demonstrates how LLM-based automation, adaptive assessment, and multimedia-enhanced content generation can be combined to support personalized learning experiences. The proposed system has potential applications in higher education, self-directed learning, and digital content creation, offering a scalable and intelligent solution for modern educational needs.
