生成式人工智能赋能软件工程课程教学改革研究

Research on Teaching Reform of Software Engineering Courses Empowered by Generative Artificial Intelligence

摘要: 针对软件工程课程教学中理论与实践结合不够紧密、项目实践过程指导不足、课程评价方式相对单一等问题，本文探索生成式人工智能赋能软件工程课程教学改革的实践路径。研究构建了“项目驱动–AI辅助–教师引导–多元评价”的教学模式，并以“学生成绩管理系统”为项目载体，将DeepSeek、Kimi和PlantUML工具融入课前预习、课中案例分析、课后反馈改进、项目全过程支持和过程评价环节。教学实践表明，该模式能够增强学生学习参与度，促进学生在需求分析、系统设计、编码测试和项目文档撰写等任务中提升工程实践能力，并推动课程评价由结果评价向过程评价转变。研究可为人工智能背景下软件工程课程教学改革提供参考。

Abstract: In response to problems in software engineering course teaching, such as the insufficient integration of theory and practice, inadequate guidance during project practice, and relatively single course evaluation methods, this paper explores a practical path for teaching reform of software engineering courses empowered by generative artificial intelligence. The study constructs a teaching model of “project-driven, AI-assisted, teacher-guided, and multi-dimensional evaluation.” Taking the “Student Grade Management System” as the project carrier, DeepSeek, Kimi, and PlantUML are integrated into pre-class preparation, in-class case analysis, after-class feedback and improvement, whole-process project support, and process evaluation. Teaching practice shows that this model can enhance students’ learning participation, promote the improvement of their engineering practice ability in tasks such as requirements analysis, system design, coding and testing, and project document writing, and facilitate the transformation of course evaluation from result-oriented evaluation to process-oriented evaluation. This study can provide a reference for teaching reform of software engineering courses in the context of artificial intelligence.