PFI:BIC - Development, Deployment and Evaluation of an Intelligent Service System for Personalized Early Literacy Learning Using Mobile Devices

PFI:BIC - 使用移动设备进行个性化早期识字学习的智能服务系统的开发、部署和评估

• 批准号: 1717362
• 负责人: Cynthia Breazeal
• 金额: $ 99.98万
• 依托单位: Massachusetts Institute of Technology
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-08-01 至 2021-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1717362&HistoricalAwards=false
• 关键词: PFI; BIC; Development; Deployment; Evaluation

Early learning is a key time to intervene for subsequent educational and economic achievement. Research shows that reading levels in early primary years can predict high school graduation rates. Effective Pre-K educational intervention programs are recognized as an effective return on investment (ROI) for society via improved employment opportunities, lower incarceration rates, and decreased need for social services. The scalability and ease-of-deployment of smart devices and other digital learning technologies has transformative potential to increase Kindergarten readiness outcomes that is widely accessible and affordable. Especially for children who cannot access quality pre-K education. However, modern digital early educational tools do not possess AI and machine learning algorithms for continuous assessment and modeling of children?s literacy and and oral language skills, nor are they capable of personalization to meet the diverse needs of individual children. The goal of this proposed research is to bridge this gap between the need for personalized interventions for early childhood learning with the ubiquity and affordability of mobile devices. The research team proposes to develop, deploy, and evaluate a system capable of automatic and continuous assessment, personalized content adaptation, and AI-powered educational mentorship to deliver high quality, scalable, and affordable early literacy interventions to young children.The proposed research will develop a smart system for personalized early learning on mobile devices, by developing three innovative technology components: 1. Embedded student assessment algorithms, that can estimate reading skill from diverse, multimodal data sources including audio, facial expression, and real-time gameplay/usage data. 2. Personalized mentoring algorithms that adapt educational content to individual patterns of behavior of app usage to optimize for individual learning outcomes. 3. Affect-aware content adaptation and personalization algorithms to optimize engagement and learning. These technologies shall be embedded within two novel interactive educational apps. The first is for storytelling and vocabulary development. The second is for reading.  Both use general assessment data, real-time audio, and facial expression recognition to estimate performance and to optimize motivation, engagement, and learning. These technologies will be embedded within industry partner Curious Learning's mobile software platform, which currently delivers early learning software for children to promote early literacy readiness of US kindergarten children. The project will evaluate the ability of these technologies to assess reading skill, foster vocabulary and oral language, and to improve engagement and learning by personalizing the Curious Learning system's educational content to children.

早期学习是为以后的教育和经济成就进行干预的关键时期。研究表明，小学早期的阅读水平可以预测高中毕业率。有效的学前教育干预计划被认为是通过改善就业机会，降低监禁率和减少对社会服务的需求为社会提供有效的投资回报（ROI）。智能设备和其他数字学习技术的可扩展性和易于部署性具有变革潜力，可以提高幼儿园的准备成果，这些成果可以广泛获得和负担得起。然而，现代数字化早教工具并不具备人工智能和机器学习算法来对儿童进行持续评估和建模？儿童的读写能力和口头语言能力都很差，也不能个性化，以满足个别儿童的不同需要。这项拟议研究的目标是弥合儿童早期学习个性化干预的需求与移动的设备的普及和可负担性之间的差距。研究团队建议开发、部署和评估一个能够自动和持续评估、个性化内容调整和人工智能驱动的教育指导的系统，为幼儿提供高质量、可扩展和负担得起的早期识字干预。拟议的研究将开发一个智能系统，用于移动的设备上的个性化早期学习，通过开发三个创新技术组件：1.嵌入式学生评估算法，可从多种多模态数据源（包括音频、面部表情和实时游戏/使用数据）评估阅读技能。2.个性化的指导算法，使教育内容适应应用程序使用的个人行为模式，以优化个人学习成果。3.情感感知的内容调整和个性化算法，以优化参与和学习。这些技术将被嵌入到两个新的互动教育应用程序中。第一个是讲故事和词汇发展。第二个是阅读。 两者都使用一般评估数据、实时音频和面部表情识别来评估绩效，并优化激励、参与和学习。这些技术将嵌入行业合作伙伴Curious Learning的移动的软件平台，该平台目前为儿童提供早期学习软件，以促进美国幼儿园儿童的早期识字准备。该项目将评估这些技术的能力，以评估阅读技能，促进词汇和口语，并通过个性化的好奇学习系统的教育内容，以提高参与和学习的儿童。

项目成果

Towards Transferrable Personalized Student Models in Educational Games

教育游戏中可迁移的个性化学生模型

• 发表时间: 2021
• 期刊: Proceedings of the 20th International Conference on Autonomous Agents and MultiAgent Systems
• 作者: Spaulding, S.

Impact of Interaction Context on the Student Affect-Learning Relationship in Child-Robot Interaction

儿童与机器人交互中交互情境对学生情感-学习关系的影响

• DOI: 10.1145/3319502.3374822
• 发表时间: 2020
• 期刊: ACM/IEEE International Conference on Human-Robot Interaction
• 作者: Chen, Huili;Park, Hae Won;Zhang, Xiajie;Breazeal, Cynthia
• 通讯作者: Breazeal, Cynthia

Frustratingly Easy Personalization for Real-time Affect Interpretation of Facial Expression

• DOI: 10.1109/acii.2019.8925515
• 发表时间: 2019-09
• 期刊: 2019 8th International Conference on Affective Computing and Intelligent Interaction (ACII)
• 作者: Samuel Spaulding;C. Breazeal