QUARTERMASTER: Query and User-based Abductive Reporting Tool Enabling Responsive Multimodal Analysis of Simulated Technological Enterprise Records

Aptima, Inc., with partners Central Washington University and nou Systems, Inc., is developing QUARTERMASTER to help DoD analysts focus on questions of interest rather than the means of answering them. QUARTERMASTER (Query and User-based Abductive Reporting Tool Enabling Responsive Multimodal Analysis of Simulated Technological Enterprise Records), is an interactive cognitive assistant meant to broaden the bench of analytical personnel. QUARTERMASTER allows analysts to express domain-specific questions in natural language to retrieve quantitative data of interest, eliminating the need to know data structure or database query languages. Moreover, QUARTERMASTER can augment these data with analyses of its own to aid human reasoning. It is also capable of explaining how it obtained a particular outcome. QUARTERMASTER is conceptually novel in several ways: (1) it uses natural language processing for querying simulation tasks that are inherently quantitative; (2) it is domain-specific; (3) it efficiently bridges both ontologies and databases; and (4) it uses causal reasoning to infer user intent and meaning as well as the root causes of results returned from database queries. Ultimately, QUARTERMASTER will empower analysts at all expertise levels to query their datasets naturally and thus contribute to faster and more complete analyses of simulations. Approved for Public Release |21-MDA-10789 (21 Apr 21)