From Data to Discovery: New PRISM Software Reveals How Research is Connected

Published October 15, 2026

Left: Data extraction and formatting pipeline for graph visualization. Right: The 13 most-published researchers and the 32 papers linking them (≥4 shared coauthors each).

CI Compass students working with MagLab data developed PRISM, a processing pipeline used to generate a scientific knowledge graph containing information about instruments, datasets, researchers, and publications. In the future, detailed knowledge graphs will be used to track the impact of the MagLab’s research and for new artificial intelligence-based workflows that will enable new scientific discoveries.

What is the development?

The MagLab’s Center for FAIR and Open Science worked with students from the CI Compass Fellowship Program to develop new software called PRISM (Provenance-Rich Index of Scientific Metadata). This system extracts metadata (data that describes other data)1 from multiple sources (such as data acquired using MagLab instruments and journal articles written by the scientists who collected it) and structures it into a knowledge graph. A knowledge graph contains nodes of information linked together with relationships that give them meaning, e.g. a researcher node linked to a publication node to indicate authorship.

Why is this important?

Scientific knowledge graphs link the people, processes, and products of research together into detailed and dynamic database of information. They allow researchers to track research products throughout their lifecycle so they can understand where and how they are used. They can also be used by artificial intelligence agents with reasoning capabilities for the discovery of new knowledge.

Who did the research?

David S. Butcher¹, Diya Adhikari², Veronika Saiadian³

¹National High Magnetic Field Laboratory, ²Caldwell University, ³University of Southern California

Why did they need the MagLab?

As a national laboratory serving a large user community and multiple disciplines of research, MagLab instrumentation and datasets are linked to a large number of researchers and publications, providing an ideal test case for demonstrating the utility of scientific knowledge graphs. While the initial implementation concerns a single user facility at the MagLab, the PRISM metadata extraction, formatting, and visualization pipeline can be used to expand the scientific knowledge graph to include more researchers, datasets, and publications, increasing its utility for both human and AI-driven scientific discovery.