Semantic Analysis Technologies for the Identification of Dual Use Research of Concern (STIR IIB)

The goal of this project is to develop the Semantic Analysis Technologies for the Identification of Dual Use Research of Concern (STIR) capability.  STIR processes scientific documents using semantic technologies and inference algorithms to identify potential Dual Use Research of Concern (DURC).  The focus is on 15 high consequence pathogens and toxins and 7 experimental categories that have been identified as DURC.  STIR evaluates multiple semantic processing and inference approaches to generate the optimal DURC classification.  STIR is optimized based on adaptability, precision, recall, time and effort required of subject matter experts, and ease of use.  Both deep semantics (extracting relationships at the molecular and cellular level, identifying biochemical entities, species, hosts, and their relationships) and shallow semantics are explored and their performance is analyzed.  Fusion of the results of different DURC identification models is performed to optimize overall STIR capability.  The currently proposed effort is motivated by the need to generalize and extend the results of STIR Phase I and II efforts.  The proposed effort will extend and generalize semantic extraction, information processing, and knowledge-based inferencing to support DURC and bioscience compliance.  It will also pursue technology transfer applications.