Data Platform Integrated Knowledge Enhancement (DataPIKE)

Leveraging the advanced hybrid knowledge graph technology that serves as the foundation of two prior non-SBIR prototypes, Cougaar Software, Inc. (CSI) will build a distributed, resilient and secure hybrid knowledge graph capability that will reside onĀand extend each C3T Tactical Data Platform node in the tactical domain.ĀThis knowledge-level capability is called the Data Platform Integrated Knowledge Enhancement (DataPIKE).ĀThis direct to Phase II will create a demonstrable prototype of the DataPIKE concept designed and developed to augment the existing C3T Tactical Data Platform and demonstrate the power of knowledge level situational reasoning as a capability augmentation to the current data platform capabilities.ĀThe purpose of this node is to add the æknowledge-levelÆ capabilities to the data platform while preserving all of its current capabilities.ĀDataPIKE will focus on building and demonstrating unique knowledge-level capabilities to enhance and empower the C3T Tactical Data Platform to include: (1) Transformation of Data Platform Data and Information into Knowledge; (2) Derivation and Monitoring of the Operating Situation; (3) Knowledge-level Situational Visualization; (4) Semantic-level Query and Assertion Validation; (5) Temporal and Spatial Pattern Analysis and Prediction; (6) Readiness, Capabilities, and State Estimation and Projection; and (7) Plan Management, Plan Development and Plan Monitoring.Ā The DataPIKE capability is composed of a core hybrid knowledge graph engine, which holds the knowledge graph and provides the core knowledge graph services such as spatial, temporal, state, persistence, and others.ĀThe knowledge graph engine is loaded with a variety of reasoners, which are small units of computation operating on the graph in response to create, update, and delete events for specific graph nodes and graph relationships.Ā Outside the knowledge graph is a set of cognitive agents, referred to as librarians, which are responsible for maintain the consistency, completeness and accuracy of the knowledge graph as new data flows in and operations are performed.Ā These librarians are privilege agents which have higher level access and control capabilities than other agents in the system, empowering them to do their curation and maintenance activities. Each node also has a set of functional and system cognitive agents which compose the majority of the æbusiness logicÆ for the capability.Ā These agents are typically built around a single core functionĀand use the platform services and knowledge graph as the basis for performing that function.ĀIn many cases, the functional or system agent also hosts a REST/JSON service, which exposes its capabilities as a micro-service to the rest of the network.Ā With the appropriate credentials, non-agent systems can utilize the capabilities, and the knowledge graph understanding, for use in other systems and capabilities.