SCC-CIVIC-PG Track B: Assessing the Feasibility of Systematizing Human-AI Teaming to Improve Community Resilience

SCC-CIVIC-PG 轨道 B：评估系统化人类与人工智能协作以提高社区复原力的可行性

• 批准号: 2043522
• 负责人: Keri Stephens
• 金额: $ 4.99万
• 依托单位: University of Texas at Austin
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-01-15 至 2022-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2043522&HistoricalAwards=false
• 关键词: SCC; CIVIC; PG; Track; Assessing

Emergency managers need access to the right data to effectively and efficiently plan for and respond to disasters. Social media offers a data source that is increasingly relevant for disaster management, but emergency response organizations typically lack capacity to monitor and mine social media data at scale. One remedy is to pair human volunteers, who label relevant information, with computers to train and update artificial intelligence (AI) systems for scalable monitoring. Including local volunteers in the process is important because they are uniquely capable of identifying locally-relevant images, text, and conversations that reflect their communities. Yet, we currently have no mechanism to systematically pair these human volunteer/AI-systems with emergency management organizations. Therefore, the fundamental issue this project investigates is the feasibility of leveraging the strengths of local members of a Community Emergency Response Team (CERT) with AI—called human-AI teaming—to bridge this gap. The unique CIVIC aspect of this project is to leverage existing collaborations with a CERT organization to assess the feasibility. The long-term vision is to develop a sustainable, replicable, and empirically informed framework for integrating CERT volunteers into the automated processing of social media data using an AI-based system. The project supports education and diversity by providing research experiences to diverse students, as well as training CERT volunteers in social media and human-AI teaming. Findings can help emergency managers better train their volunteers who comb through social media using understandings of the built environment to help machines see new patterns in data. Hence, this project supports NSF's mission to promote the progress of science and advance the nation's health, prosperity, and welfare by demonstrating the value of leveraging local CERT volunteers, in partnership with emergency managers, to generate disaster situation awareness. The goal of this planning grant is to analyze existing human-AI teaming disaster data and involve civic partners in focus groups to better understand the attitudes and beliefs of CERT volunteers, emergency managers, key governmental organizations, and non-governmental organizations. This project will develop deep knowledge of digital volunteer teams, how they work, how to motivate them, and how to have them support the objectives of emergency managers. Thus, we advance theory around volunteer teaming in the technology space and human-in-the-loop protocols. This project provides meaningful ways for more citizens to participate in disaster planning and response, and develops a training curriculum for CERT volunteers who work with social media data in an effort to build sustainable volunteer efforts.This project is in response to Track B - CIVIC Innovation Challenge - Resilience to Natural Disasters a collaboration with NSF and the Department of Homeland Security.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.

应急管理人员需要访问正确的数据，以有效地有效地计划和应对灾难。社交媒体提供了与灾难管理越来越相关的数据源，但是应急组织通常缺乏按大规模监控和挖掘社交媒体数据的能力。一种补救措施是将标记相关信息标记的人类志愿者与用于训练和更新人工智能（AI）系统进行可扩展监视的计算机。在此过程中包括本地志愿者很重要，因为他们具有独特的能力来识别反映其社区的本地图像，文本和对话。但是，我们目前尚无将这些人类志愿者/人工智能系统与应急管理组织系统配对的机制。因此，该项目调查的基本问题是利用社区应急小组（CERT）的当地成员的优势，以AI（称为人类的组合）来弥合这一差距。该项目的独特公民方面是利用与证书组织的现有合作来评估可行性。长期的愿景是开发一个可持续，可复制和经验知情的框架，以使用基于AI的系统将志愿者整合到社交媒体数据的自动处理中。该项目通过为潜水员的学生以及社交媒体和人类AI团队的培训志愿者提供研究经验来支持教育和多样性。调查结果可以帮助应急管理人员更好地培训他们的志愿者，他们使用对建筑环境的理解来帮助机器看到数据中的新模式。因此，该项目支持NSF促进科学进步的使命，并通过证明利用当地证书志愿者与紧急情况管理人员合作的价值来促进国家的健康，繁荣和福利，以提高灾难状况意识。这项计划赠款的目标是分析现有的人类AI团队灾难数据，并让公民合作伙伴参与焦点小组，以更好地了解CERT志愿者，急诊经理，主要政府组织和非政府组织的出勤率和信念。该项目将发展对数字志愿团队的深入了解，他们的工作方式，如何激励他们以及如何使他们支持紧急经理的目标。这是我们围绕技术领域和人类在循环协议中的志愿者团队的理论提高了理论。该项目为更多公民参与灾难规划和反应提供了有意义的方式，并为与社交媒体数据合作的证书志愿者开发了培训课程，以努力建立可持续的志愿者努力。该项目是为了跟踪B-公民创新挑战 - 与自然灾害的恢复性 - 与自然灾害的恢复能力，与NSF和HOMELAND SECRATING a DISTISS FARTICE nsf REDATITY a DISTISTIONS NSF REVERITY NSFREATIONS。使用基金会的智力优点和更广泛的影响评估标准进行评估。

项目成果

Implementation of a multidisciplinary COVID-19 social media capability in uncertain times.

在不确定时期实施多学科的 COVID-19 社交媒体功能。

• 发表时间: 2021
• 期刊: International Association of Emergency Managers
• 作者: Peterson, S.

Human-AI teaming for COVID-19 response: A practice & research collaboration case study

人类与人工智能合作应对 COVID-19：一种实践

• 发表时间: 2022
• 期刊: Proceedings of the 19th International ISCRAM Conference
• 作者: Hughes, A.;Stephens, K. K.;Peterson, S.;Purohit, H.;Harris, A. G.;Senarath, Y.;Jarvis, S. A.;Montagnolo, C. E.;Nader, K.
• 通讯作者: Nader, K.