Improving Programming Skills of Engineering Students at Historically Black Colleges and Universities Using AI-enhanced Personalized Adaptive Learning Tools

该项目旨在通过建立在线人工智能(AI)增强的个性化自适应学习(PAL)工具来服务于国家利益，以提高历史上的黑人学院和大学(HBCU)的工程学学生的编程技能。编程技能是工程教育中各种模拟的必备技能。随着新冠肺炎的流行从根本上改变了课程交付方式，在线教学工具变得不可或缺。该项目团队计划开发一系列用于编程教学和学习的在线学习模块，通过深度强化学习推荐PAL路径，并使用人工智能技术构建智能编程助手(SPA)来改善学生的学习。拟议的在线AI增强型PAL工具有可能将PAL教学法与强化学习等人工智能技术相结合，以探索如何推荐PAL学习路径并为工科学生生成适当的参考源代码。项目团队打算完成三项主要任务。首先是开发具有开放应用编程接口的基本在线工具，这些工具能够与新的人工智能算法相连接，以改善学生的学习。第二，使用深度强化学习推荐PAL路径，以最大限度地提高学习者在适当难度水平下的编程学习参与度。第三，也是最后一点，是通过基于深度学习的语言模型来设计SPA，以生成参考代码来帮助学习者进行编程活动。建议的工具有可能帮助教授其他工程学课程。该团队打算在各种网络研讨会、开放源码库和教育会议上传播他们的项目成果。拟议的工具将与其他HBCU和机构共享。NSF IUSE：EDU计划支持研究和开发项目，以提高所有学生的STEM教育的有效性。通过其参与的学生学习跟踪，该计划支持有前途的实践和工具的创建、探索和实施。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This project aims to serve the national interest by building online Artificial Intelligence (AI) enhanced personalized adaptive learning (PAL) tools to improve engineering students' programming skills at Historically Black Colleges and Universities (HBCUs). Programming skills are imperative for various simulations in engineering education. As the COVID-19 pandemic has fundamentally changed course delivery methods, online teaching tools have become indispensable. The project team plans to develop a series of online learning modules for programming teaching and learning, recommend PAL paths via deep reinforcement learning, and build a smart programming assistant (SPA) using AI techniques to improve student learning. The proposed online AI enhanced PAL tools have the potential to integrate a PAL pedagogy with AI techniques such as reinforcement learning to explore how to recommend PAL learning paths and produce appropriate reference source codes for engineering students. The project team intends to work on three primary tasks. First is to develop basic online tools with open application programming interfaces that are able to connect with novel AI algorithms to improve student learning. Second is to recommend PAL paths using deep reinforcement learning to maximize learners' engagement in programming learning at proper difficulty levels. Third, and finally, is to design an SPA via deep learning-based language models to generate reference codes to help learners with programming activities. The proposed tools have the potential to help to teach other engineering courses. The team intends to disseminate their project results at various webinars, open-source repositories, and education conferences. The proposed tools would be shared with other HBCUs and institutions. The NSF IUSE: EDU Program supports research and development projects to improve the effectiveness of STEM education for all students. Through its Engaged Student Learning track, the program supports the creation, exploration, and implementation of promising practices and tools.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.