Abstract:

This paper explores the transformative role of ChatGPT in the teaching mode centered on knowledge concepts. As a language generation model, ChatGPT is capable of in-depth language comprehension and innovative combinations by mining the symbiotic relationships between words through massive language data learning. However, in the field of education, ChatGPT faces limitations such as over-reliance on training data, weak logical reasoning ability, and limited ability to handle new scenarios. To overcome these limitations and enhance the accuracy and relevance of ChatGPT’s generated content, this paper proposes an organic combination of ChatGPT with the organization of teaching resources centered on knowledge concepts, and improve ChatGPT by creating structure diagrams of knowledge concepts. Additionally, several specific and feasible ways to assist teachers and students using ChatGPT are also proposed. Finally, this paper discusses how to combine the prompt research paradigm with the teaching mode centered on knowledge concepts to help ChatGPT establish a “knowledge system”. This will enable ChatGPT to become a language generation model driven by both data and knowledge, providing more intelligent and personalized services in the education field, and promoting its development and transformation.

Key words: ChatGPT, AIGC; large language model, knowledge concept, teaching resource organization, personalized learning