Semantical Machine Understanding Navy SBIR FY2006.2 Sol No.: Navy SBIR FY2006.2 Topic No.: N06-153 Topic Title: Semantical Machine Understanding Proposal No.: N062-153-1005 Firm: Quantum Intelligence, Inc. 3375 Scott Blvd Suite 100 Santa Clara, California 95054-3111 Contact: Ying Zhao Phone: (408) 980-0090 Web Site: www.quantumii.com Abstract: Defense Transformation has changed warfighting tactics, requiring quick reaction, team-based mobile force operations in discrete events. Joint, Coalition, Non-Government and Volunteer Organizations need to analyze open-source, uncertain, conflicting, partial, non-official data. More powerful information analysis tools are needed that can quickly extract meaning, intent and specified semantic content from large volumes of unstructured multilingual text and represent it mathematically. We propose to develop an innovative semantical machine learning, understanding and search architecture to meet the needs of Semantical Machine Understanding. Our innovation is to integrate an agent social network and text mining within a reinforcement learning framework, therefore, it is cognitive, human-in-loop and real-time. It is an infrastructure allowing incorporating human interactions in the loop to gradually enhance machine understanding. Real-time and collaborative machine learning with initial innovative text mining is to provide the ultimately practical approach for semantical machine understanding and search. If successful, this paradigm would be much powerful than a static text analysis tool and would be readily applied as the search technology in a planned operational test environment at a Navy or Defense Intelligence Analysis venue and identify events, relationships and trends to enable interoperable knowledge sharing and intelligence analysis across joint and coalition forces. Benefits: The proposed approach draws many diversified technologies for the total benefits of Semantical Machine Understanding and Search. It is set up like a P2P capable of access real-time and dynamic information. It is keyword-based, capable of sophisticated global ranking like a centralized search engine. It integrates innovative reinforcement learning, social network and text mining technologies to achieve a cognitive level of machine understanding. Meaning providers also use the scheme to learn and strengthen their capabilities with real-time new experience. If successful, the result technology will dramatically increase the capability of distributed cooperative machine understanding. Private-sector applications would include any information analysis situation that involves high data volume and quick response requirements. This would include state and local emergency support teams for crisis action planning and humanitarian aid response. Business applications, such as corporate knowledge management or textual research would also be a target. The product from this project can be applied for a wide range of DOD applications, for example, to apply the result in a distributed collaborative coordination of mobile ad-hoc networks services in sensor networks, law reinforcement, disaster management, emergency response for homeland security, e.g. chemical, bio, nuclear detection and monitoring, traffic observation, detection and management, industrial control systems, portable and mobile site monitors. Commercial potential for the technology would include any collaborative or team problem solving situation where it is necessary to maintain a team consensus on an issue or product in a dynamically changing environment. The corporate buyers of extended enterprises can use the product for access and locate knowledge services across an entire enterprise including partners and vendors. Small business owners can use the product to build storefronts with service location and discovery capabilities of each other across an entire community of knowledge services including coordinating search and understanding like computers, cell phones, printers, and other electronic devices (include light switches, thermostats, alarms, televisions). Return