Battelle, a nonprofit global research and development organization, has entered into an agreement with the Intelligence Advanced Research Projects Activity to develop a predictive technology intended for monitoring epidemic-related threats.
The researchers are set to incorporate the TechAware Search Continuous Automated Scanning for Technology Transformation methodology into an application after completing the project’s pilot phase, which included the creation and validation of the prototype, Battelle said Tuesday.
The CASTT system is designed to track text data from different streams through natural language processing. Using the information, it builds a graph neural network and forecasts significant behavioral changes that are pandemic-related.
“This project shows significant promise for the use of graph neural networks to predict real-world, rare events using high-volume text data,” said Allen Chen, lead data scientist at Battelle.
Analysts will move on to testing the system’s effectiveness as part of the next phase and Chinese information sources will be included in the database.