Prof. Suzanne Barber of UT ECE, along with co-PIs Prof. Lauren Meyers of the Division of Statistics and Scientific Computation and Andy Ellington of the Department of Chemistry and Biochemistry, have been awarded a grant by the Department of Defense Defense Threat Reduction Agency (DTRA) for their work titled "Surety BioEvent App." The grant is worth nearly $3 Million over three years. 

The work will develop a BioSurveillance Virtual Environment (BSVE) Service called the Surety BioEvent App Trust Filter. This filter will describe the trust confidence for a specific data source. The trust tuple has six dimensions: Identity, Expertise, Reputation, Experience, Authority, Separation. This trust tuple will be used by a data-driven goal-oriented optimization AI approach (multi-objective sequential optimization algorithm) that will match data source surety to specific surveillance goals and constraints. This approach was used to optimize Texas' flu surveillance system by combining traditional (physician reporting) with NextGen (Google Flu Trends) to maximize early detection and statewide situational awareness for both seasonal and pandemic influenza. Using this approach, they discovered that over-report of flu from urban areas impeded surveillance and that Google Flu Trends was useful for seasonal flu but was misleading for pandemic flu. They propose that this approach will enable selecting data sources that work best for specific surveillance goals.

This proposal offers a very innovative approach to determine the relevance and value (trustworthiness) of various data sources, including traditional, OSI and SM. In addition, this effort has a novel approach to use an AI-based method to optimize data streams based on surveillance goals and constraints set by the BSVE User. This effort will enable DTRA to enhance its BSV offerings with a general mechanism to assess utility of data streams for its BSV applications.