Collaborative Research: EAGER: AI-Assisted Just-in-Time Scaffolding Framework for Exploring Modern Computer Design

合作研究：EAGER：用于探索现代计算机设计的人工智能辅助即时脚手架框架

• 批准号: 2327971
• 负责人: Mohammad Alian
• 金额: $ 20万
• 依托单位: University of Kansas Center for Research Inc
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-08-15 至 2025-07-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2327971&HistoricalAwards=false
• 关键词: Collaborative; Research; EAGER; AI; Assisted

There is a need to train skilled computer architects to design innovative computer hardware. Software-based simulation is the backbone of computer system design and development. Such tools are also widely used for teaching computer architecture concepts. Currently, the simulators used for educational purposes have a steep learning curve, are not interesting for beginners, and are error-prone. Therefore, these simulators are mostly used by experienced researchers. This project introduces a novel framework and technology called Scaffolded AI-driven Learning Simulation (SAILS). SAILS enables an interactive and supportive computer architecture learning platform and offers design exercises covering different learning modes and difficulty levels. In the development phase, SAILS will be used by instructors at the University of Kansas and Florida International University to teach introductory and advanced computer architecture courses to about 400 undergraduate and graduate computer science and engineering students every year. Once SAILS is fully developed, it will serve as a framework to teach computer architecture in several US institutions.SAILS implements a novel AI-driven paradigm for reducing the learning curve of computer architecture simulators in educational settings. SAILS implements a front end that reduces the complexity of simulating simple to advanced systems for students with various backgrounds. SAILS back-end seamlessly connects to a state-of-the-art computer architecture simulator and provides just-in-time personalized assistance to the users. The assistance is provided by a centralized AI model trained by individual users’ and team data and global users’ experience with the framework. SAILS integrates the faded scaffolding approach to provide appropriate levels of support to individual learners and teams to maximize their learning. SAILS’s easy-to-use graphical user interface, engaging learning activities, and personalized scaffolding support a broad and diverse student population, including female and underrepresented minority students, in the computer architecture and design field.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.

需要培训熟练的计算机架构师来设计创新的计算机硬件，这种工具也广泛用于计算机体系结构概念的教学。学习曲线陡峭，对初学者来说不感兴趣，而且容易出错，因此，这些模拟器主要由经验丰富的研究人员使用。该项目引入了一种称为支架式人工智能驱动学习模拟（SAILS）的新颖框架和技术。 SAILS 提供了一个交互式和支持性的计算机架构学习平台，并提供涵盖不同学习模式和难度级别的设计练习。在开发阶段，SAILS 将被堪萨斯大学和佛罗里达国际大学的教师用来教授入门和高级计算机架构课程。每年向大约 400 名计算机科学与工程专业的本科生和研究生提供服务。一旦 SAILS 完全开发出来，它将成为美国多家机构教授计算机架构的框架。SAILS 实现了一种新颖的人工智能驱动范例，以缩短计算机的学习曲线。建筑学SAILS 实现了一个前端，可以降低为具有不同背景的学生模拟简单到高级系统的复杂性，SAILS 后端无缝连接到最先进的计算机体系结构模拟器，并提供即时的服务。该帮助由由个人用户和团队数据训练的集中式 AI 模型提供，SAILS 集成了褪色的脚手架方法，为个人学习者提供适当水平的支持。 SAILS 易于使用的图形用户界面、引人入胜的学习活动和个性化支架为计算机体系结构和设计领域的广泛且多样化的学生群体（包括女性和代表性不足的少数族裔学生）提供支持。法定使命，并通过使用基金会的智力优点和更广泛的影响审查标准进行评估，被认为值得支持。