Working Group D

Agenda
	Working Groups

Working Group D	Readying Data for AI/ML through Ontologies, Ontology Standards, and Best Practices
	Wednesday- February 26, 2025 1:00 PM – 5:00 PM PT
Description	The need for semantic consistency across vast and highly segregated volumes of data remains a critical issue for the Department of Defense and Intelligence communities. Semantic consistency is a shared understanding of what a group of terms (vocabulary) and relationships between these terms mean across groups of developers, subject matter experts and throughout an enterprise. Without this shared meaning, advanced analytics, artificial intelligence, machine learning (AI/ML) and desired automation provided by these technologies can provide incorrect results or fail due to faulty data interpretation. “Our intelligence communities have access to more data than ever before. Analytic processes generate even more data that can be leveraged for further analysis. Our Intelligence and Defense communities continue to seek out methodologies for harnessing of this ever-increasing volume of data, including the use of automation, but to date, the speed and potential of this objective is critically impaired due to data issues that include inconsistencies in data format, inconsistencies in how data is represented in software, lack of alternate names for terms (labels), and a lack of a formal language that specifies how data relates to other data (formal logic). Formal logic allows new facts and relationships to be inferred based on existing facts and relationships positively impacting analytics and AI/ML. The use of ontologies provides a path for reconciling these issues, but only if the ontologies are developed according to standards and best practices to encourage future utility, scalability, and interoperability. Adherence to these principles encourage reuse and enable cost reduction for development and maintenance of the ontologies.” This working group applies to any missions, including space and ground systems/applications, even beyond the DoD and IC, where ontologies are being used or considered; in addition, this working group applies to those who need to “ready” their data for AI/ML. This Working Group builds upon the successful DIOWG Evening Session, held at GSAW 2024, with 35+ participants. Proposed Format: The Working Group will begin with ensuring awareness and a common understanding of ontologies. That is, an overview of ontologies will be provided, including what they are and why they are of value to “ready” data for AI/ML. The overview will include some examples of how they could be applied to space and ground systems/applications, and other missions. Next, the Working Group, led by Dr. Limbaugh, Dr. Beverley, and Mr. Jensen, will move to an interactive discussion with the Working Group participants of some of the current “hard problems” for applying ontologies across the community and building enduring, interoperable data architectures that adopt ontology best practices. Finally, the Working Group leads, along with additional ontology and content SMEs, will close with a Q&A session, set in a panel format.
Leads	David Limbaugh, Ph. D, National Center for Ontological Research, John Beverley, Ph. D University of Buffalo, and Mark Jensen, U.S. Customs and Border Protection
Biographies	Dr. David Limbaugh’s bio currently unavailable Dr. John Beverley is an Assistant Professor at the University of Buffalo and Co-Director of the National Center for Ontological Research. His research areas are formal logic and knowledge representation concerning a range of topics such as information, maritime awareness, and biomedicine, among others. Dr. Beverley has a B.S. in Philosophy from North Carolina State University, an M.A. in Philosophy from SUNY University at Buffalo, and a Ph.D. in Philosophy from Northwestern University.  Mr. Mark Jensen is a Data Scientist with U.S. Customs and Border Protection and formerly a Senior Ontologist at CUBRC Inc. He has been a developer for the Common Core Ontologies (CCO) since 2015 and is the chair of the Governance and Developers Group for CCO. His current areas of interest are standards for ontologies, automating validations for semantic content, competency metrics, and how use cases and application requirements can determine the fidelity of ontology-based data models. Mr. Jensen has a B.A. in Cognitive Science-Symbolic Systems and an M.A. in Philosophy-Applied Ontology, both from SUNY University at Buffalo.