Building on previous internal investments and leveraging ongoing advancements in semantic technologies, PNNL implemented a formal reasoning framework and applied it to a specific challenge in nuclear nonproliferation. The Semantic Nonproliferation Analysis Platform (SNAP) was developed as a preliminary graphical user interface to demonstrate the potential power of the underlying semantic technologies to analyze and explore facts and relationships relating to the nuclear fuel cycle (NFC). In developing this proof of concept prototype, the utility and relevancy of semantic technologies to the Office of Defense Nuclear Nonproliferation Research and Development (DNN R&D) has been better understood.

Final Report for the NA22 Simulations, Algorithm and Modeling (SAM) Ontology Assessment Team's efforts from FY09-FY11. The Ontology Assessment Team began in May 2009 and concluded in September 2011. During this two-year time frame, the Ontology Assessment team had two objectives: (1) Assessing the utility of knowledge representation and semantic technologies for addressing nuclear nonproliferation challenges; and (2) Developing ontological support tools that would provide a framework for integrating across the Simulation, Algorithm and Modeling (SAM) program. The SAM Program was going through a large assessment and strategic planning effort during this time and as a result, the relative importance of these two objectives changed, altering the focus of the Ontology Assessment Team. In the end, the team conducted an assessment of the state of art, created an annotated bibliography, and developed a series of ontological support tools, demonstrations and presentations. A total of more than 35 individuals from 12 different research institutions participated in the Ontology Assessment Team. These included subject matter experts in several nuclear nonproliferation-related domains as well as experts in semantic technologies. Despite the diverse backgrounds and perspectives, the Ontology Assessment team functioned very well together and aspects could serve as a model for future inter-laboratory collaborations and working groups. While the team encountered several challenges and learned many lessons along the way, the Ontology Assessment effort was ultimately a success that led to several multi-lab research projects and opened up a new area of scientific exploration within the Office of Nuclear Nonproliferation and Verification.

This paper presents the multi-layer, data-driven advanced reasoning tool (M-DART), a proof-of-principle decision support tool for improved power system operation. M-DART will cross-correlate and examine different data sources to assess anomalies, infer root causes, and anneal data into actionable information. By performing higher-level reasoning “triage” of diverse data sources, M-DART focuses on early detection of emerging power system events and identifies highest priority actions for the human decision maker. M-DART represents a significant advancement over today's grid monitoring technologies that apply offline analyses to derive model-based guidelines for online real-time operations and use isolated data processing mechanisms focusing on individual data domains. The development of the M-DART will bridge these gaps by reasoning about results obtained from multiple data sources that are enabled by the smart grid infrastructure. This hybrid approach integrates a knowledge base that is trained offline but tuned online to capture model-based relationships while revealing complex causal relationships among data from different domains.

We describe the design and development of an advanced reasoning framework employing semantic technologies, organized within a hierarchy of computational reasoning agents that interpret domain specific information. Designed based on an inspirational metaphor of the pattern recognition functions performed by the human neocortex, the CHAMPION reasoning framework represents a new computational modeling approach that derives invariant knowledge representations through memory-prediction belief propagation processes that are driven by formal ontological language specification and semantic technologies. The CHAMPION framework shows promise for enhancing complex decision making in diverse problem domains including cyber security, nonproliferation and energy consumption analysis.

The insider threat ranks among the most pressing cyber-security challenges that threaten government and industry information infrastructures. To date, no systematic methods have been developed that provide a complete and effective approach to prevent data leakage, espionage, and sabotage. Current practice is forensic in nature, relegating to the analyst the bulk of the responsibility to monitor, analyze, and correlate an overwhelming amount of data. We describe a predictive modeling framework that integrates a diverse set of data sources from the cyber domain, as well as inferred psychological/motivational factors that may underlie malicious insider exploits. This comprehensive threat assessment approach provides automated support for the detection of high-risk behavioral "triggers" to help focus the analyst's attention and inform the analysis. Designed to be domain-independent, the system may be applied to many different threat and warning analysis/sense-making problems.

We describe the modeling of concepts taken from social and behavioral science theories, specifically Social Movement Theory, with OWL ontologies. One goal of the study was to provide analysts with a way to understand the potential contribution of social movement research to their tasks without interrupting workflow. Another goal was to facilitate knowledge transfer at the point of need between busy analysts and the latest research. Instead of providing mechanisms for obtaining consensus on debated concepts, we used different namespaces for each author and record bibliographic information in another ontology. We also describe a framework for knowledge discovery supported by the ontologies.

Semantic Web applications require robust and accurate annotation tools that are capable of automating the assignment of ontological classes to words in naturally occurring text (ontological annotation). Most current ontologies do not include rich lexical databases and are therefore not easily integrated with word sense disambiguation algorithms that are needed to automate ontological annotation. WordNet provides a potentially ideal solution to this problem as it offers a highly structured lexical conceptual representation that has been extensively used to develop word sense disambiguation algorithms. However, WordNet has not been designed as an ontology, and while it can be easily turned into one, the result of doing this would present users with serious practical limitations due to the great number of concepts (synonym sets) it contains. Moreover, mapping WordNet to an existing ontology may be difficult and requires substantial labor. We propose to overcome these limitations by developing an analytical platform that (1) provides a WordNet-based ontology offering a manageable and yet comprehensive set of concept classes, (2) leverages the lexical richness of WordNet to give an extensive characterization of concept class in terms of lexical instances, and (3) integrates a class recognition algorithm that automates the assignment of concept classes to words in naturally occurring text. The ensuing framework makes available an ontological annotation platform that can be effectively integrated with intelligence analysis systems to facilitate evidence marshaling and sustain the creation and validation of inference models.

The Universal Parsing Agent (UPA) is a document analysis and transformation program that supports massive scale conversion of information into forms suitable for the semantic web. UPA provides reusable tools to analyze text documents; identify and extract important information elements; enhance text with semantically descriptive tags; and output the information that is needed in the format and structure that is needed.

Software environments for intelligence analysis need to leverage scenario-based analytical processes to help analysts achieve a more comprehensive view of competing hypotheses in their risk assessments. While current solutions offer some of the tools and functionality required for such advancement, there still is no integrated set of capabilities that addresses the extraction and manipulation of scenario content from unstructured intelligence data sources. We propose to fill this gap through the development of a visual interactive environment for event analytics that integrates text mining, discourse analysis and visualization capabilities to support scenario manipulation and generation processes.

Semantic Web applications require robust and accurate annotation tools that are capable of automating the assignment of ontological classes to words in naturally occurring text (ontological annotation). Most current ontologies do not include rich lexical databases and are therefore not easily integrated with word sense disambiguation algorithms that are needed to automate ontological annotation. WordNet provides a potentially ideal solution to this problem as it offers a highly structured lexical conceptual representation that has been extensively used to develop word sense disambiguation algorithms. However, WordNet has not been designed as an ontology, and while it can be easily turned into one, the result of doing this would present users with serious practical limitations due to the great number of concepts (synonym sets) it contains. Moreover, mapping WordNet to an existing ontology may be difficult and requires substantial labor. We propose to overcome these limitations by developing an analytical platform that (1) provides a WordNet-based ontology offering a manageable and yet comprehensive set of concept classes, (2) leverages the lexical richness of WordNet to give an extensive characterization of concept class in terms of lexical instances, and (3) integrates a class recognition algorithm that automates the assignment of concept classes to words in naturally occurring text. The ensuing framework makes available an ontological annotation platform that can be effectively integrated with intelligence analysis systems to facilitate evidence marshaling and sustain the creation and validation of inference models.