NSF Corner: Directorate for Technology, Innovation, and Partnerships

To strengthen the connection between the statistical community and National Science Foundation, we continue the series introduced in the May 2023 issue that poses questions to NSF program officers and awardees. If you have questions or comments for the program officers, send them to ASA Director of Science Policy Steve Pierson.

This month’s program officer is Jemin George of the NSF Directorate for Technology, Innovation, and Partnerships. The awardee responses are from Yuzhou Chen of Temple University.

Program Director

Jemin George joined the National Science Foundation in January 2022 as a detailee from the US Army Combat Capability Development Command Army Research Laboratory. He serves as a program director for the Innovation and Technology Ecosystems Division within the Directorate for Technology, Innovation, and Partnerships and works closely with the Convergence Accelerator Program to establish and evaluate strategic initiatives and policies. He also serves as liaison for activities connected with the US Department of Defense and other agencies and sectors, including private industry. George joined the Army Research Laboratory in 2010 and has been responsible for identifying technical gaps and effectively formulating and executing basic and applied research in several areas of importance to the Army. He has served as the Army’s collaborative link to more than 30 academic/industry partners while conducting Army-relevant disruptive foundational research, which has been documented in more than 120 peer-reviewed technical articles.

What Innovation and Technology Ecosystems programs are most relevant to the statistical and data science community?

For the statistical and data science community, the Innovation and Technology Ecosystems Division programs that stand out as highly relevant for the statistical and data science community are the following:

• Pathways to Enable Open-Source Ecosystems: This program serves the statistical and data science community by fostering the development of open-source tools critical for data analytics, machine learning, and statistical modeling. Its commitment to building and nurturing open-source ecosystems offers statisticians and data scientists a platform for contributing to and using cutting-edge tools and libraries that facilitate collaborative research and innovation. It also provides statisticians and data scientists the opportunity to lead projects to develop new open-source statistical methods, machine learning algorithms, and data visualization tools. Finally, the program emphasizes creating sustainable ecosystems, encouraging statisticians and data scientists to engage in community-driven development efforts and ensuring the longevity and relevance of open-source tools.
• Building the Prototype Open Knowledge Network: This program is particularly relevant to those in data science and statistics due to its focus on developing interconnected knowledge graphs, which are instrumental in organizing and querying large-scale, complex data sets. It aligns with the core competencies of data scientists and statisticians in semantic data integration, big data analytics, and the development of artificial intelligence–driven insights from vast data sources.

  The Building the Prototype Open Knowledge Network provides unique opportunities for collaboration and innovation. The creation of domain-specific knowledge graphs presents an opportunity for statisticians and data scientists to apply their expertise in data modeling and analysis to contribute to the foundational infrastructure of the Open Knowledge Network. By participating in this program, the statistical and data science community can lead the way in leveraging knowledge graphs for advanced data fusion techniques, improving the accuracy and efficiency of predictive modeling and data-driven decision-making processes.

• Responsible Design, Development, and Deployment of Technologies: Although broader in scope, this program is critical to the data science community for its focus on ethical AI and responsible technology development. It aligns with the increasing demand for transparency, fairness, and accountability in AI systems and machine learning models, areas in which statisticians and data scientists are inherently involved.

  Data scientists have a pivotal role in shaping the principles and practices around responsible AI, from developing bias mitigation techniques to ensuring privacy-preserving data analysis. By engaging in the Responsible Design, Development, and Deployment of Technologies program, statisticians and data scientists can collaborate with experts from various fields to ensure technological innovations are developed and deployed in a manner that is beneficial and equitable to society.

• Accelerating Research Translation: This program emphasizes the translation of fundamental research into practical applications, a core aspiration of the data science field. It supports projects that develop data-driven technologies, algorithms, and applications with potential for real-world impact. Statisticians and data scientists in academia can leverage this program to bridge the gap between theoretical research and societal applications, turning innovative concepts into deployable solutions. The Accelerating Research Translation program’s focus on training graduate students and postdoctoral researchers in translational research offers a pathway for emerging data scientists to gain valuable skills in entrepreneurship and applied research.

In summary, these programs provide a rich landscape for statisticians and data scientists to lead and contribute to projects that not only advance the frontiers of their field but also ensure the responsible development and deployment of technology for societal benefit. Through collaborative efforts, open-source contributions, ethical technology development, and research translation, the statistical and data science community has numerous opportunities to significantly affect both science and society.

What kind of research proposals does the Innovation and Technology Ecosystems Division typically fund?

The division focuses on funding research proposals geared toward the translation of scientific discoveries and innovations into societal and economic benefits. The nature of research proposals can be diverse, but they share common themes of innovation, collaboration, and potential for impact. Here’s a closer look at the types of research proposals the division is inclined to fund:

• Use-Inspired Research: The division looks for proposals directly inspired by societal needs and offering solutions to critical challenges through innovative technology or methodologies. Projects that can bridge the gap between fundamental research and practical applications, demonstrating a clear path to implementation and adoption in real-world scenarios.
• Translational Research: Projects that focus on moving scientific and engineering discoveries from the laboratory to the marketplace or society at large. This includes developing technologies, products, or processes with potential commercial applications or societal impact. The division supports efforts across various stages of technology readiness, from early-stage concept validation to prototype development and pilot deployment.
• Multidisciplinary and Collaborative Efforts: The division values research proposals that bring together expertise from different disciplines, sectors, and stakeholders to address complex problems. This includes partnerships between academia, industry, non-profits, and government agencies.
• Open-Source and Knowledge-Sharing Initiatives: The division is interested in projects that facilitate sharing knowledge, data, and technologies, promoting accessibility and collaboration within the scientific community and beyond. This includes efforts to build open knowledge networks and infrastructures that enable data integration and interoperability.
• Responsible and Inclusive Innovation: The division seeks projects that address issues of fairness, privacy, security, and trust, ensuring technological advancements are accessible, inclusive, and beneficial for all segments of society.
• Education and Workforce Development: The division is interested in projects that contribute to educating and training the future workforce in areas critical to national competitiveness and innovation, including initiatives that integrate research experiences, entrepreneurship, and translational research skills into STEM education at all levels.

The division’s funding priorities reflect a comprehensive approach to innovation, emphasizing the importance of advancing technology for the public good, fostering economic growth, and addressing societal challenges. Proposals that align with these goals, demonstrate a clear vision for impact, and incorporate collaborative, multidisciplinary efforts stand out in the funding landscape.

Does the Innovation and Technology Ecosystems program involve principal investigators whose primary appointment is in a statistics department?

Innovation and Technology Ecosystems programs welcome and involve PIs from an array of disciplinary backgrounds, including those whose primary appointment is in statistics departments. This engagement is reflective of the division’s commitment to fostering interdisciplinary research and development projects that leverage the strengths of diverse fields to address complex challenges and drive innovation.

Statistics PIs are crucial to the division’s mission due to their expertise in data analysis, statistical modeling, and quantitative methodologies. These skills are indispensable in projects focused on big data, machine learning, artificial intelligence, and computational modeling, which are increasingly central to solving contemporary societal and technological challenges.

Innovation and Technology Ecosystems programs not only fund research but also emphasize workshops, educational initiatives, and the development of tools statisticians are uniquely positioned to lead or significantly contribute to. These activities foster a broader understanding of data science across disciplines and promote the responsible use of statistical methodologies in technology development.

The division encourages collaboration across disciplines, sectors, and institutions. Statisticians in Innovation and Technology Ecosystems–funded projects often work alongside engineers, computer scientists, social scientists, and industry professionals, underscoring the interdisciplinary nature of modern research endeavors.

Statisticians, with their primary appointment in statistics departments, are not only eligible but are integral to the success of Innovation and Technology Ecosystems–funded projects. Their expertise in statistical theory and methods—combined with the division’s focus on innovative, translational, and interdisciplinary research—makes for a synergistic relationship that advances both the statistics field and broader goals of technological innovation and societal benefit. Through Innovation and Technology Ecosystems, statisticians have unique opportunities to apply their knowledge to real-world problems, collaborate across disciplines, and contribute to the development of ethical and responsible technologies.

What are the few key items the statistical/data science community should know about the Innovation and Technology Ecosystems programs?

Innovation and Technology Ecosystems programs offer a unique opportunity to contribute to and shape the future of interdisciplinary research and innovation. Relevant overarching themes are the following:

• Emphasis on Interdisciplinary Collaboration: Innovation and Technology Ecosystems strongly encourages projects that foster collaboration across different disciplines, integrating statistical and data science expertise with other scientific and engineering domains. This interdisciplinary approach is designed to tackle complex research questions and societal challenges more effectively, providing statisticians and data scientists with the opportunity to apply their skills in diverse contexts.
• Focus on Translational Impact: Innovation and Technology Ecosystems programs are dedicated to translating fundamental research findings into practical applications that benefit society. For statisticians and data scientists, this means an emphasis on projects that not only advance theoretical understanding but also have clear pathways to implementation and real-world impact. It’s about moving beyond the data to deliver solutions that matter.
• Open Science and Accessibility: A commitment to open science underpins the division’s mission, encouraging the development of open-source tools, methodologies, and data-sharing practices. This principle aligns with the data science community’s efforts to enhance reproducibility, transparency, and collaborative innovation through accessible data and software.
• Ethical and Responsible Innovation: Recognizing the profound implications of emerging technologies, the division promotes research that incorporates ethical considerations, privacy, fairness, and inclusivity from the outset. For the statistical and data science community, this underscores the importance of developing and applying algorithms and models that are not only technically sound but also socially responsible.
• Support for Workforce Development: Innovation and Technology Ecosystems places a strong emphasis on education and workforce development, aiming to prepare the next generation of researchers, practitioners, and innovators. This focus includes initiatives to integrate research experiences, entrepreneurial skills, and translational research capabilities into STEM education, directly benefiting statisticians and data scientists at all career stages.
• Contribution to Innovation Ecosystems: Finally, the division aims to nurture and grow innovation ecosystems that bridge academia, industry, and government. For statisticians and data scientists, participating in Innovation and Technology Ecosystems programs offers a chance to be at the forefront of creating and applying knowledge in ways that drive technological advancement, economic growth, and societal well-being.

In summary, the division’s overarching themes highlight the critical role of the statistical and data science community in advancing interdisciplinary research, fostering open and ethical innovation, and contributing to societal and economic benefits. Engaging with Innovation and Technology Ecosystems programs offers unique opportunities to lead impactful research, collaborate across disciplines, and train future leaders in data science and beyond.

Awardee

Yuzhou Chen is an assistant professor in the department of computer and information sciences at Temple University. He earned his PhD in statistics at Southern Methodist University in 2021. His research focuses on geometric machine learning, topological data analysis, and knowledge discovery and is supported by National Science Foundation Directorate for Technology, Innovation, and Partnerships and Division of Mathematical Sciences grants, as well as NASA Advanced Information Systems Technology grants.

Chen and his collaborators were recently awarded funding from the Directorate for Technology, Innovation, and Partnerships for “Develop Dynamic, REsponsive, Adaptive, and Multifaceted Knowledge Graphs (DREAM-KG) to Address Homelessness with Explainable AI,” which aims to introduce the emerging concept of knowledge graphs to a comprehensive statistical analysis of the social, economic, and political factors that contribute to homelessness while triaging existing resources to support homeless people.

Chen’s primary collaborators include Chiu Tan of Temple University, Huanmei Wu of Temple University, Ying Ding of The University of Texas at Austin, Karin Eyrich-Garg of Temple University, and Omar Martinez of the University of Central Florida. Their grant was funded at $1.5 million and will be used for the development of AI-assisted models and the associated statistical uncertainty quantification, data collection, knowledge graph construction and storage, training of large language models, and implementation of the proposed approaches in the form of publicly available software.

Tell us more about the project, including its goal.

Homelessness is a growing problem of the utmost societal importance in many countries around the world, exacerbated by the COVID-19 pandemic and economic downturn. It is particularly acute among those with underprivileged socioeconomic status. The project aims to create a novel systematic way to improve the reliability, efficiency, and sustainability of the analyzing homelessness by leveraging emerging AI models, knowledge graphs, and innovative statistical approaches for uncertainty quantification of multi-modal data. Students working on the proposed research will acquire an excellent orientation to cross-disciplinary research training at the interface of data science, statistics, and biomedical and social sciences.

We will introduce the emerging concept of knowledge graphs to a comprehensive statistical analysis of the social, economic, and political factors that contribute to homelessness, while triaging existing resources to support homeless people. Knowledge graphs are founded on the principle of applying a graph-based abstraction to data and are now broadly deployed in scenarios that require integrating and extracting value from large-scale multimodal data. The primary users of the system will be frontline case workers, nonprofit organizations, federal agencies, and, hopefully, homeless people. More generally, while knowledge graphs remain largely unexplored from the statistical perspective, I believe synergy between formal statistical approaches and the wealth of information contained in knowledge graphs can open numerous new directions that will benefit statisticians, various domain experts, and society.

Please describe your approach to the Directorate for Technology, Innovation and Partnerships.

I am a statistician by training with an interdisciplinary flavor (my BA is in applied math, my MSc and PhD are in statistics, my postdoc is in electrical engineering, and now I am with the computer science department). I have always been interested in the intersection of artificial intelligence, machine learning, and statistics. Through communicating with researchers from different communities, I found my expertise can be perfectly linked with many use-inspired challenges—from health informatics to power grid resiliency to wildfire predictive analytics. I enjoy interdisciplinary research, and I have conducted numerous projects fusing statistics and AI.

Regarding our DREAM-KG project supported by the NSF Directorate for Technology, Innovation, and Partnerships, my team and I reached out to many experts from different institutions—academic to nonprofit organizations—trying to ensure our ideas would indeed push the envelope of the current approaches to tackle homelessness and, at the same time, be relevant to this Directorate for Technology, Innovation, and Partnerships call on knowledge graphs. Our project, similar to many other Directorate for Technology, Innovation, and Partnerships awards, has a strong focus on use-inspired and translational aspects. So, for example, we spent a lot of time identifying proper stakeholders, understanding their needs, and making sure the products to be developed would be implemented in the real world.

What advice do you have for others applying for NSF funding?

First, read the solicitation of call for proposal in detail. Second, participate in the webinar hosted by the program directors of the target call for proposal or make a one-on-one meeting with program directors and discuss whether the proposed ideas fit into the call. This is enormously useful—you can receive many invaluable suggestions. Third, check prior projects funded by the target program and learn more about the successful proposals. Moreover, ensure the proposal is easy to read/follow and can be understood by reviewers from different communities (beyond mathematics and statistics or your specific research domain). Fourth, do not be afraid to get out of your comfort zone and ask others for feedback—from program directors to potential future collaborators in industry and nonprofit organizations. I think the last point is particularly important for Directorate for Technology, Innovation, and Partnerships funding.