arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.23830cs.HC
收藏

CodeExemplar:面向生成式人工智能时代的入门编程示例式支架

CodeExemplar: Example-Based Scaffolding for Introductory Programming in the GenAI Era

Faculty of Arts and Science, Kyushu University(九州大学文理学院)

Boxuan Ma, Shinichi Konomi

AI总结:

本文提出基于示例的支架方法,通过生成与目标任务推理模式相符但情境不同的示例,支持类比迁移并减少抄袭,设计了二维分类体系和指导原则,并开发了CodeExemplar原型系统。

摘要:

生成式人工智能(GenAI)能够以最小的努力生成可工作的代码,这在入门编程中产生了矛盾:学生需要及时帮助,但直接解决方案可能引发抄袭并阻碍推理。为此,我们提出了基于示例的支架方法,其中GenAI提供与目标任务底层推理模式相符但情境不同的支架示例,以支持类比迁移并减少抄袭。我们贡献了一个二维分类体系、设计指南和CodeExemplar原型系统,该系统集成了自动评分任务,并通过课堂试点和教师访谈进行了初步形成性反馈。

英文摘要:

Generative AI (GenAI) can generate working code with minimal effort, creating a tension in introductory programming: students need timely help, yet direct solutions invite copying and can short-circuit reasoning. To address this, we propose example-based scaffolding, where GenAI provides scaffold examples that match a target task's underlying reasoning pattern but differ in contexts to support analogical transfer while reducing copying. We contribute a two-dimensional taxonomy, design guidelines, and CodeExemplar, a prototype integrated with auto-graded tasks, with initial formative feedback from a classroom pilot and instructor interviews.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑