Emerging applications and instrumental systems produce massive streaming data every day. It is essential that the big streaming data be quickly and accurately processed and analyzed to extract useful information for timely decision making. Despite substantial work being proposed for streaming data processing and analysis, e.g., SAMOA, Jubatus, etc., many shortcomings have not been well addressed. InfoBeyond advocates Context-aware Machine Learning Middleware for Real-time Distributed Streaming Big Data Analysis (COMAND) to address the challenges of distributed massive streaming data analysis. COMAND includes machine learning (ML) based algorithms that are designed to facilitate its implementation and integration with the existing infrastructures and algorithms. The operation is implemented via a two-stage operational architecture for efficient and accurate processing of massive streaming data. We will first design and develop a middleware architecture that enables efficient and scalable capabilities in distributed environments. Next, we will analyze and develop three key algorithms, namely, Distributed Streaming Data Clustering (DISDC), Distributed Optimal Context-Aware Data Classification (DOCDC), and Real-time Publish/Subscribe Data Model Update (RPSDMU) for the middleware. The system is wrapped by APIs and Adapters for facilitating the implementation and integration of COMAND with the existing systems.

Missile defense faces new challenges for national defense due to a spectrum of increasingly capable air and missile delivery systems and the means to counter them. Complex and integrated attacks are possible using a set of guided rockets, artillery, and mortars; unmanned aerial vehicles (UAVs); a range of land attack and anti-ship ballistic and cruise missiles; increasingly maneuverable ballistic missile reentry vehicles; hypersonic glide vehicles; anti-satellite weapons; and active air and missile defense interceptors. Consequently, understanding and classifying inbound sophisticated weapons quickly provides critical information necessary to perform optimal defense through precision assignment and targeting of the highest priority objects. However, current defense systems are incapable of dealing with the contextual uncertainty imposed by target proximity features (e.g., the time before hit), capability features (e.g., envelope), human judgment (e.g., target's intent), and relevant social features as a whole to yield optimal response during the full stages of missile defense.

DB&EL (Threat Classification and Reasoning by Dynamic Bayesian Network, Description Logic, and Intuitionistic Fuzzy Sets) is a dynamic reasoning technology to provide threat classification refinement with advanced contextual deep learning capabilities. It takes all the sensing data, text/messages, and human judgments into account for optimal defense decision-making through the full life cycle of a threat. Advanced contextual reasoning is achieved by leveraging Bayesian theory, dynamic graphs, Description Logic (DL), text mining, and Fuzzy technologies. Dynamic Bayesian Network (DBN) can derive threat classification or degrees of threat confidence that allow probabilistic event reasoning (e.g., cause and consequence probability) while fusing the spatial and temporal dependent features of observations. Influence Diagram (ID) with DL is empowered for sophisticated contextual reasoning, such as the likelihood of a consequence to hold of a threat giving the context, and the optimistic and pessimistic outcome of a responding strategy. It then performs a risk assessment to sort the degree of threats if multiple threats appear such that defending resources can be optimally allocated.

DB&EL aligns well with missile defense applications such as the ground midcourse defense system, offering value propositions from aspects of classification refinement and strategy optimization. DB&EL can be beneficial for clinical reasoning and diagnosis of Depression, Bipolar Disorder, Schizophrenia, Dementia, Post-traumatic Stress Disorder, etc., where these diseases exhibit symptom uncertainty. It can also be used for terrorist identification, cybersecurity risk assessment, disaster management, and other applications where contextual information should be incorporated for reasoning under uncertainty.

Emerging applications and instrumental systems produce massive streaming data every day. It is essential that the big streaming data be quickly and accurately processed and analyzed to extract useful information for timely decision making. Despite substantial work being proposed for streaming data processing and analysis, e.g., SAMOA, Jubatus, etc., many shortcomings have not been well addressed. InfoBeyond advocates Context-aware Machine Learning Middleware for Real-time Distributed Streaming Big Data Analysis (COMAND) to address the challenges of distributed massive streaming data analysis. COMAND includes machine learning (ML) based algorithms that are designed to facilitate its implementation and integration with the existing infrastructures and algorithms. The operation is implemented via a two-stage operational architecture for efficient and accurate processing of massive streaming data. We will first design and develop a middleware architecture that enables efficient and scalable capabilities in distributed environments. Next, we will analyze and develop three key algorithms, namely, Distributed Streaming Data Clustering (DISDC), Distributed Optimal Context-Aware Data Classification (DOCDC), and Real-time Publish/Subscribe Data Model Update (RPSDMU) for the middleware. The system is wrapped by APIs and Adapters for facilitating the implementation and integration of COMAND with the existing systems.

Missile defense faces new challenges for national defense due to a spectrum of increasingly capable air and missile delivery systems and the means to counter them. Complex and integrated attacks are possible using a set of guided rockets, artillery, and mortars; unmanned aerial vehicles (UAVs); a range of land attack and anti-ship ballistic and cruise missiles; increasingly maneuverable ballistic missile reentry vehicles; hypersonic glide vehicles; anti-satellite weapons; and active air and missile defense interceptors. Consequently, understanding and classifying inbound sophisticated weapons quickly provides critical information necessary to perform optimal defense through precision assignment and targeting of the highest priority objects. However, current defense systems are incapable of dealing with the contextual uncertainty imposed by target proximity features (e.g., the time before hit), capability features (e.g., envelope), human judgment (e.g., target's intent), and relevant social features as a whole to yield optimal response during the full stages of missile defense.

DB&EL (Threat Classification and Reasoning by Dynamic Bayesian Network, Description Logic, and Intuitionistic Fuzzy Sets) is a dynamic reasoning technology to provide threat classification refinement with advanced contextual deep learning capabilities. It takes all the sensing data, text/messages, and human judgments into account for optimal defense decision-making through the full life cycle of a threat. Advanced contextual reasoning is achieved by leveraging Bayesian theory, dynamic graphs, Description Logic (DL), text mining, and Fuzzy technologies. Dynamic Bayesian Network (DBN) can derive threat classification or degrees of threat confidence that allow probabilistic event reasoning (e.g., cause and consequence probability) while fusing the spatial and temporal dependent features of observations. Influence Diagram (ID) with DL is empowered for sophisticated contextual reasoning, such as the likelihood of a consequence to hold of a threat giving the context, and the optimistic and pessimistic outcome of a responding strategy. It then performs a risk assessment to sort the degree of threats if multiple threats appear such that defending resources can be optimally allocated.

DB&EL aligns well with missile defense applications such as the ground midcourse defense system, offering value propositions from aspects of classification refinement and strategy optimization. DB&EL can be beneficial for clinical reasoning and diagnosis of Depression, Bipolar Disorder, Schizophrenia, Dementia, Post-traumatic Stress Disorder, etc., where these diseases exhibit symptom uncertainty. It can also be used for terrorist identification, cybersecurity risk assessment, disaster management, and other applications where contextual information should be incorporated for reasoning under uncertainty.

The U.S.Navy was in search for a real-time graph embedding tool for analyzing huge graphs (millions of nodes and billions of edges) from diverse sources. However, current approaches cannot provide dynamic and scalable graph analytics to show the military value of tactical data. In this project, InfoBeyond advocates EStreaming (Embedding & Streaming) for scalable and efficient graph streaming. EStreaming promotes big data streaming technology where unsupervised and semi-supervised machine learning algorithms can be conducted over the streaming platform.

It can split a huge graph into small subgraphs such that distributed graph embedding can be conducted in parallel among a set of processors. Meanwhile, the graph embedding can be effectively merged and visualized. Considering the diversity of Navy applications, EStreaming is an open platform that can implement many graph embedding algorithms such as IsoMap, LLE, Laplacian eigenmaps, and graph factorizations. We have demonstrated the implementation of Distributed LINE (DLINE) and Distributed MVE (DMVE). Compared to other algorithms, classification of DLINE can be conducted from the internal relations and the similarity among the persons in the solider, enemy, and other social networks. Differently, DMVE can be used for analyzing spatial-temporal data such as for ISR sensor networks, e.g., Navy ISR geographic traffic monitoring.