摘要
当前,软件工程课程实践化教学面临着“软件案例”备课设计耗时、“软件案例”动态演示困难、“项目实践”开源重构受限等挑战。以ChatGPT为代表的自然语言生成大模型已初步具备了生成软件开发与演化相关预测性内容的能力,在需求分析、软件设计、代码生成、软件测试等核心开发活动中展现出极强能力。分析当前软件工程类课程实践化教学方法及主要局限,并介绍大模型驱动的人机协作智能化软件开发与运行新范式,最后探讨这一新范式在软件工程实践化教学中的潜在应用前景。
Currently,the practice-oriented teaching of software engineering courses faces several challenges,including the time-consuming preparation of"software case"materials,difficulties in dynamically demonstrating"software cases,"and limitations in open-source reconstruction for"project practice."Large language models,represented by ChatGPT,have initially demonstrated the ability to generate predictive content related to software development and evolution,showcasing strong capabilities in core development activities such as requirements analysis,software design,code generation,and software testing.This paper first analyzes the current methods and main limitations of practice-oriented teaching in software engineering courses,then introduces a new paradigm of human-machine collaborative intelligent software development and operation driven by large language models,and finally discusses the potential applications of this new paradigm in the practice-oriented teaching of software engineering.
作者
杨硕
毛新军
许凯
唐建平
肖志群
王鹏
张琪
YANG Shuo;MAO Xinjun;XU Kai;TANG Jianping;XIAO Zhiqun;WANG Peng;ZHANG Qi(College of Systems Engineering,National University of Defense Technology;College of Computer Science,National University of Defense Technology,Changsha 410073,China)
出处
《软件导刊》
2025年第11期173-177,共5页
Software Guide
基金
国防科技大学青年自主创新科学基金项目(ZK2023-36)
国防科技大学本科和任职教育教学研究课题(U2024013)。
关键词
软件工程实践教学
软件案例
大模型
软件开发新范式
practice-oriented teaching of software engineering
software cases
large models
new paradigm of software development
