The U.S. National Science Foundation (NSF) has launched the NSF Open Knowledge Network (NSF OKN), a national open-data infrastructure designed to connect knowledge graphs and make data more accessible, interoperable, and machine-readable for research, government, and other applications.
The network is publicly available at okn.us and currently links 43 knowledge graphs containing tens of billions of connected facts spanning areas including health, the environment, justice, manufacturing, and national security.
NSF OKN provides a structured and verifiable data layer that can complement artificial intelligence systems by providing access to source information, attribution, and structured knowledge. The network is designed to give AI systems an interoperable and auditable representation of information that can be used alongside conventional AI models.
The initiative addresses challenges associated with datasets that are distributed across separate systems and are not always interoperable or prepared for machine-based applications. By connecting federal and public datasets, NSF OKN is intended to make information more discoverable and usable across different fields.
NSF has identified AI-ready datasets and next-generation scientific data infrastructure as areas of investment as part of its broader efforts to support AI-enabled scientific discovery. The Open Knowledge Network represents one implementation of this approach by providing infrastructure for connecting and querying data across domains.
The NSF OKN originated through the Prototype Open Knowledge Network program, which began in 2023 with $26.7 million in funding for 18 research teams. The program received core support from the National Institutes of Health (NIH), National Aeronautics and Space Administration (NASA), National Institute of Justice (NIJ), National Oceanic and Atmospheric Administration (NOAA), and U.S. Geological Survey (USGS).
The participating teams initially developed knowledge graphs for individual fields and connected them through a common technical infrastructure. This created a federated network that links the individual graphs rather than maintaining them as separate projects.
The network subsequently expanded beyond the original 18 projects as federal agencies contributed additional resources and data. NIH provided $2.5 million in supplemental funding to integrate 10 of its knowledge graph initiatives, while NASA, USGS, and NOAA contributed additional data products, including a coastal-zone digital twin and NASA's Earth-observing knowledge graph.
The initiative now includes 43 knowledge graphs, with participation from more than 12 federal agencies and more than 90 cross-sector partnerships.
One of the network's capabilities is cross-graph querying, which allows information from different datasets to be analyzed together. For example, connections between supply-chain and manufacturing data can be used to examine cybersecurity exposure across more than 7,000 semiconductor products from 247 companies. Similarly, combining contamination and soil-productivity data can support analysis of relationships between environmental conditions and agricultural productivity.
The network is intended for use beyond the research community. Potential applications include helping caseworkers identify housing-related information, supporting communities in preparing for floods, and allowing companies to examine software supply chains for potential vulnerabilities.
Click here to read the original press release.
