AI Institute for Inclusive Intelligent Technologies for Education (INVITE)

AI 普惠智能教育技术研究所 (INVITE)

• 批准号: 2229612
• 负责人: H Chad Lane
• 金额: $ 1999.87万
• 依托单位: University of Illinois at Urbana-Champaign
• 依托单位国家: 美国
• 项目类别: Cooperative Agreement
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-06-01 至 2028-05-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2229612&HistoricalAwards=false
• 关键词: AI; Institute; Inclusive; Intelligent; Technologies

The Institute for Inclusive and Intelligent Technologies for Education (INVITE) seeks to fundamentally reframe how educational technologies interact with learners by developing artificial intelligence (AI) tools and approaches to support three crucial noncognitive skills known to underlie effective learning: persistence, academic resilience, and collaboration. This new generation of systems will be radically more responsive to learner needs, behaviors, and development and be designed to support the whole learner, beyond discipline-focused achievement. Use-inspired research will focus on how children communicate STEM content, how they learn to persist through challenging work, and how teachers support and promote noncognitive skill development. The resultant AI-based tools will be integrated into classrooms to empower teachers to support learners in more developmentally appropriate ways. This work will generate a rich set of data documenting learners’ interactions with educational technologies, each other, and teachers, allowing researchers to study learner growth over time and across different STEM activities. Research and outreach activities will draw from the INVITE K-12 partner network reaching up to 96,000 learners across 24 school districts and nonprofits spanning 8 states. The Institute will offer inclusive programs to support diverse students’ participation in research experiences, undergraduate courses in AI in education, and professional development programs for teachers.Institute research will pursue foundational AI advances in robust and fair machine learning, learner modeling, and natural language understanding to enable assessment and modeling of noncognitive skill development over time and across domains. It will revolve around three interconnected strands: (1) Collect, analyze, and share novel datasets for fair and robust machine learning and natural language understanding; (2) Build novel, robust methods for understanding learner behaviors and persistent, integrated learner models that incorporate assessments of noncognitive skills; (3) Develop new inclusive STEM learning environments that provide natural and adaptive interaction with socially-aware pedagogical agents. Interpretable generative models fit to real data and simulated learners will enable new discoveries and hypotheses about human learning. Use-inspired research will advance the science of noncognitive skill acquisition during STEM learning and uncover relevant contextual aspects of learning historically overlooked by AI systems. The institute will serve as a nexus for building capacity for research, education, and broadening participation in the intersection of AI and Education for All, serving a wide array of stakeholders. Specifically, the Institute will (1) produce a database of multimodal datasets for use by other researchers, (2) provide open source tools and opportunities to develop knowledge about the use, control, and impact of innovative AI-enabled education systems, and (3) actively build a diverse workforce of future scientists and engineers to design, implement, and deploy the next generation of AI-enabled Education for All systems. The National Center for Education Research at the Institute of Education Sciences of the US Department of Education is partnering with NSF to provide funding for the Institute.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.

包容性和智能教育技术研究所 (INVITE) 致力于从根本上重新构建教育技术如何与通过人工智能 (AI) 工具和方法开发的学习者互动，以支持已知的有效学习基础的三种关键非认知技能：坚持、学术弹性和学习能力。新一代系统将从根本上更加响应学习者的需求、行为和发展，并旨在支持整个学习者，超越以学科为中心的成就，以使用为导向的研究将重点关注儿童如何交流 STEM 内容，以及他们如何进行学习。学会坚持通过具有挑战性的工作，以及教师如何支持和促进非认知技能的发展，由此产生的基于人工智能的工具将被整合到课堂中，使教师能够以更适合发展的方式支持学习者。这项工作将生成一组丰富的数据，记录学习者的互动。与教育技术、彼此和教师一起，使研究人员能够研究学习者随着时间的推移和不同 STEM 活动的成长。研究和外展活动将从 INVITE K-12 合作伙伴网络中获取，覆盖 24 个学区的多达 96,000 名学习者。该研究所将提供涵盖 8 个州的非营利性项目，以支持不同学生参与研究体验、人工智能教育本科课程以及教师专业发展项目。研究所的研究将致力于在稳健和公平的机器学习、学习者方面取得基础人工智能进步。它将围绕三个相互关联的方面进行评估和建模：（1）收集、分析和共享新颖的数据集，以实现公平和稳健的机器学习和自然语言理解。 （2）建立新颖、稳健的方法来理解学习者行为和持久、集成的学习者模型，其中纳入非认知技能的评估；(3) 开发新的包容性 STEM 学习环境，提供与具有社会意识的可解释的生成模型的自然和适应性交互。数据和模拟学习者将实现关于人类学习的新发现和假设，以使用为灵感的研究将推动 STEM 学习过程中非认知技能习得的科学，并揭示人工智能系统历史上令人惊叹的学习的相关背景。作为建设研究、教育和扩大参与人工智能和全民教育交叉点的纽带，该研究所将为广泛的利益相关者提供服务。具体来说，该研究所将 (1) 建立一个供其他研究人员使用的多模式数据集数据库。 ，（2）提供开源工具和机会来发展有关创新人工智能教育系统的使用、控制和影响的知识，以及（3）积极建立一支由未来科学家和工程师组成的多元化员工队伍，以设计、实施和部署下一代人工智能教育美国教育部教育科学研究所的国家教育研究中心正在与 NSF 合作，为该研究所提供资金。该奖项反映了 NSF 的法定使命，并通过使用评估结果被认为值得支持。基金会的智力价值和更广泛的影响审查标准。

项目成果

How Noisy is Too Noisy? The Impact of Data Noise on Multimodal Recognition of Confusion and Conflict During Collaborative Learning

• DOI: 10.1145/3577190.3614127
• 发表时间: 2023-10
• 期刊: Proceedings of the 25th International Conference on Multimodal Interaction
• 作者: Yingbo Ma;Mehmet Celepkolu;K. Boyer;Collin Lynch;E. Wiebe;Maya Israel

What makes a role model motivating for young girls? The effects of the role model’s growth versus fixed mindsets about ability and interest

是什么让榜样能够激励年轻女孩？

• DOI: 10.1016/j.jecp.2023.105775
• 发表时间: 2024
• 期刊: Journal of Experimental Child Psychology
• 影响因子: 2.6
• 作者: Gladstone, Jessica R.;Tallberg, Molly;Jaxon, Jilana;Cimpian, Andrei
• 通讯作者: Cimpian, Andrei

Computing Self-Efficacy in Undergraduate Students: A Multi-Institutional and Intersectional Analysis

计算本科生的自我效能感：多机构和交叉分析

• DOI: 10.1145/3626252.3630811
• 发表时间: 2024
• 期刊: Proceedings of the 55th ACM Technical Symposium on Computer Science Education V.1
• 作者: Ojha, Vidushi;West, Leah;Lewis, Colleen M.
• 通讯作者: Lewis, Colleen M.

IEKG: A Commonsense Knowledge Graph for Idiomatic Expressions

• DOI: 10.18653/v1/2023.emnlp-main.881
• 发表时间: 2023-12
• 期刊: ArXiv
• 作者: Ziheng Zeng;Kellen Tan Cheng;Srihari Venkat Nanniyur;Jianing Zhou;Suma Bhat

Non-compositional Expression Generation Based on Curriculum Learning and Continual Learning

• DOI: 10.18653/v1/2023.findings-emnlp.286
• 发表时间: 2023
• 作者: Jianing Zhou;Ziheng Zeng;Hongyu Gong;Suma Bhat