AIDEN: Analysis of Intelligence Data for Evidence of Networks Navy SBIR FY2010.2 Sol No.: Navy SBIR FY2010.2 Topic No.: N102-180 Topic Title: AIDEN: Analysis of Intelligence Data for Evidence of Networks Proposal No.: N102-180-0167 Firm: 21st Century Technologies Inc. 6011 West Courtyard Drive Bldg 5, Suite 300 Austin, Texas 78730 Contact: Matthew McClain Phone: (512) 682-4735 Web Site: www.21technologies.com Abstract: The goal of intelligence, surveillance, and reconnaissance (ISR) operations is to provide analysts and warfighters with situational awareness information. ISR enterprise systems store data to support these operations, but large volumes make it difficult to locate information of interest, such as key individuals. Current keyword-search approaches fall short because they fail to capture semantics. Social network analysis (SNA) techniques, including group detection and SNA metric-based classification, have been demonstrated to find valuable information in graphs, such as networks of interest and the roles of individuals in those networks. Current natural language processing (NLP) tools extract entities, themes, and relationships from unstructured text. By combining such NLP tools, information in text-based documents can be converted into a graph. The application of SNA techniques to such graphs has the potential to enable analysts and warfighters to efficiently locate actionable intelligence in ISR enterprise datastores. 21st Century Technologies (21CT) proposes Analysis of Unstructured Data for Evidence of Networks (AIDEN), a tool for locating information of interest in large, diverse data sources. Our Phase 1 work will focus on a proof-of-concept for applying SNA on graphs obtained from NLP, and identification of gaps in state-of-the-art NLP that will need to be overcome. Benefits: AIDEN will provide users with the ability to apply powerful SNA techniques to information in large volumes of unstructured data. We will leverage our existing group detection and SNA metric computation capabilities in Phase 1 and future work. We will apply advanced machine learning techniques to enable extraction of social network information from diverse unstructured data sources. AIDEN will be:  Dynamic: AIDEN will process data as is arrives to allow for up-to-date analyses  Adaptive: users will be able to train AIDEN to analyze different types of data  Flexible: users will be able to select the right types of SNA techniques to apply to find the information they need  Interactive: users will be able to review the data sources to verify results and modify the SNA graph as needed The anticipated results of AIDEN Phase I are a proof-of-concept demonstration of applying SNA to information from unstructured text and identification of gaps in current state-of-the-art NLP needed to support this type of analysis. This will identify research directions for Phase II and future work toward developing an operational system. Return