Assistant Teaching System for Computer Hardware Courses Based on Large Language Model

Recently, Large Language Models (LLMs), represented by ch1ChatGPT, have garnered significant attention in the field of education due to its impressive capabilities in text generation, comprehension, logical reasoning, and conversational abilities. We incorporate LLMs into the theoretical and experiment teaching of our Digital Logic and Computer Organization courses to enhance the teaching process. Specifically, we propose and implement an assistant teaching system consisting of a knowledge-based Question and Answer (Q &A) system and an assistant debugging and checking system. For the theoretical teaching session, the Q &A system utilizes historical Q &A records and ChatGPT to answer students’ questions. This system reduces the repetitive workload for teachers by answering similar questions, and allows students to receive answers in time. For the Field-Programmable Gate Array (FPGA)-based experiment teaching session, the assistant debugging and checking system employ debug assistance module to explain error messages for students. Furthermore, a LLM-generated code checking module assists teachers in detecting academic misconduct among students’ code submissions.