LLM-Based Course Comprehension Evaluator

Large language models (LLMs) like GPT-4 reshape intelligent tutoring systems by enabling nuanced natural language interactions. Leveraging LLMs’ capabilities, this study introduces an innovative Lesson Comprehension Evaluator, utilizing advanced Natural Language Processing (NLP) methods and Augmented Retrieval Generation (RAG) to assess course material comprehension. Through a web interface, students engage with tailored questions and receive feedback, fostering immersive learning experiences. Each response undergoes rigorous evaluation against a ground truth LLM-generated knowledge base, encompassing semantic comprehension, specificity, and correctness metrics. These evaluations provide insights into students’ course understanding, informing future pedagogical strategies. By incorporating auditory options for accessibility and gamification elements for enhanced engagement, this approach facilitates self-paced, deeper learning, fostering dynamic and enriching learning environments.