16. risk reductionadaptive feed-back Exe Model auto deploy test /demo eval COI opportunity to the direction alpha demo adaptive feed-back Each Iterative and Incremental Model Evolution follow the same structure: Exe Model auto deploy test /demo eval development
17. 7 The Vision Leading to COIN JDE Potential T-Cell C2-C-IED DCGS IED IED COIN Capabilities
18. 8 Know Adversarial Behavior and Intent Strategic Special Activities Int’l Support Strategic National Strategic Warfighter Leadership Operational Planning Timeliness of Effect Funding Tactical Recruiting TTP-R Surveillance Local Leadership Engineering Supplier Adapting Principle Success Effectiveness Overtime Production Assessment Escape Emplacement Follow-on Attack Each of the Nodes in the Network Consist of Activities which Emmitt “Signals” and Patterns that can be Detected, Analyzed and Acted Upon
20. 10 Timeliness - Combination of Event, Request and Time Driven Processing Level 1- Tactical Exploitation Level 2-Operational Exploitation Level 3/4/5- Strategic Exploitation Months To Continuous Days To Weeks Minutes To Hours
21.
22. However due to: intelligence gathering limitations, deception performed by the adversary, overlapping activities of various non-adversarial groups, etc., we need to model the activities of adversarial organizations as only partially observable.
23. The task at hand is therefore to develop a signal model that can be used to distinguish between suspicious pattern and real hostile activities. The model must be able to:
28. Prune the search of the best fit of observations to dynamic templates.
29. We will represent the signal model as an event driven transaction model, which identifies relationships between nodes in the network to describe their structure and behavior.
33. Achieve “NRT Insight”Events Signifies the Activities they Represent Aggregate Low Level Complex Event Patterns That Machines can Detect CEP is a way of distilling the information value from a number of simple events into more useful summary level complex events improving Situational Awareness and timely decision making
48. 16 Example of UI with Dynamic Template Matching Visualization 1 3 2 4 5 2: Contextual Reports 1: Data Driven Browser 3: Low level Activity Match 4: Higher level Activity Match (triggered by Lower level) 5: Notifications and Alerts The continuous dynamic template matching with human-in-the-loop control improves situational and activity awareness and promotes timely decision making based on actionable intelligence