基于生成式AI的软件工程课程教学改革研究

Research on Teaching Reform of Software Engineering Course Based on Generative AI

摘要: 针对当前高校软件工程课程存在的理论与实践脱节、工程化训练不足、个性化指导缺失、教学内容与产业技术栈脱节等突出问题，本文将生成式AI深度融入教学全流程，构建“课前智能诊断–课中师生机协同探究–课后AI精准强化–全过程智慧评价”四位一体闭环教学体系。依托AI知识图谱、智能助教、代码辅助、需求工程仿真及个性化习题与项目推送等技术手段，重构教学流程、升级实践体系、创新育人模式。该改革突出精准化、工程化、个性化与闭环化的育人特色，有效强化学生系统分析、工程实践与团队协作能力，为新时代软件工程课程高质量发展提供可复制、可推广的改革范式。

Abstract: In response to prominent problems existing in current software engineering courses in universities, such as the disconnection between theory and practice, insufficient engineering training, lack of personalized guidance, and the mismatch between teaching content and industrial technology stacks, this paper deeply integrates generative AI into the entire teaching process and constructs a four-in-one closed-loop teaching system: pre-class intelligent diagnosis, in-class teacher-student-AI collaborative inquiry, after-class AI targeted reinforcement, and whole-process intelligent evaluation. Supported by technical means including AI knowledge graphs, intelligent teaching assistants, code assistance, requirements engineering simulation, and personalized exercise and project push, the teaching process is restructured, the practical training system is upgraded, and the talent cultivation model is innovated. This reform highlights the talent cultivation characteristics of precision, engineering orientation, personalization and closed-loop management, effectively strengthens students’ abilities in system analysis, engineering practice and teamwork, and provides a replicable and promotable reform paradigm for the high-quality development of software engineering courses in the new era.