Research Summaries

Back Improve the Effectiveness of the Decision-Maker Orientation Processes in Making Sense of the Events & Phenomena Presented by the Battlespace

Fiscal Year 2019
Division Research & Sponsored Programs
Department NPS Naval Research Program
Investigator(s) Godin, Arkady A.
Murphree, James T.
Sponsor NPS Naval Research Program (Navy)
Summary The repertoire of Patterns-of-Life (POLs) Situations is derived from battlespace situations, interacting objects and subjects, during “Observation” OODA phase to answer “who, what, and how” which may be sufficient for targeting. Decision-makers during “Orientation" phase need predictive POLs Activities to answer “why”. Developing repertoire of POLs Activities speeds-up and improves effectiveness of decision-making. We propose, for the observation phase, to split up the information space into tiles covering each of the situations. Observed patterns, guided by xView data set-based (http://xviewdataset.org/) training model, will leverage Machine Learning (ML) to discover signatures for all situations. Repetitive signatures will get nominated to the POLs Situations level. This analysis breaks up the battlespace into tile-based geometrical shapes per each POL Situation. For orientation phase, ML generates POLs Activities from POLs Situations. Predictive relationships between POLs Activities and environmental METOC phenomena will be identified. The final step is to train statistical model to predict the repertoire of POLs Activities based on METOC phenomena predictions. “Events ontology” will be generated and used to link related POLs activities to improve the decision-making.

Methodology:
• Obtain non-environmental data sets and METOC phenomena predictions
• Utilize xView data set to generate repertoire of POLs Situations
• Utilize repertoire of POLs Situations to generate repertoire of POLs Activities
• Identify predictive relationships between environmental conditions and POLs Activities.
• Train ML model to predict POLs Activities from environmental phenomena forecasts
• Link related activities using “events ontology”

Deliverables: literature review on ML for OODA loop; report on ML approach and results for proposed methodology; document knowledge representation providing delineation of activity events from tiled situations.
Keywords
Publications Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal
Data Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal