Construction and Practice of Teaching Resources for University Computer Courses Based on Large AI Models

Abstract

Traditional university computer courses use textbook, static courseware, and standard laboratory environment which guarantee the stability of teaching content in case of slow update of knowledge but have become increasingly prominent with regard to knowledge lag and rigid structure under the knowledge ecology where computer technology updates at a frequency. With the ability of large AI models to generate content, it will change from pre-packaging to on-demand generation of resource construction. We analyzed the structural limitation of the traditional resource system; clarified the redefinition of the generation mechanism by large AI models raising three types of theoretical questions (interpretability, controllability, and adaptability boundaries); constructed a large AI model adaptation mechanism for computer course resource generation involving semantic modeling of the knowledge system, teaching adaptation control, and computable quality evaluation; designed a resource construction framework based on multi-stage generation process, multi-source knowledge fusion, and a dynamic evolution mechanism working together aiming to provide theoretical analysis and technical path reference for the large AI model driving construction of teaching resources for computer courses.