Distributed Collaborative Decision Support Tool
Summary:
With BMDO SBIR funding, Datamat Systems Research, Inc. (McLean, VA), has developed a software tool that can analyze large, distributed databases and transform data into useful information without moving or sharing it. This tool could help support decision-making in business, military, and government organizations. Datamat is actively marketing this software, called InferAgent™, to Federal, state, and local government agencies, U.S. armed forces, and Fortune 500 companies.
Technology Description:
With BMDO SBIR funding, Datamat Systems Research, Inc. (McLean, VA), has developed a software tool that can analyze data stored in large, distributed databases and transform it into useful information without moving or sharing it. This tool could help support decision-making in business, military, and government organizations.
Many organizations collect huge amounts of data and store it in databases at different locations. Traditionally, to extract useful information from this data, the databases must be combined together at one location. This can be difficult because database owners often have privacy restrictions placed on their data and, thus, do not want to move or share it. Datamat’s technology can solve this problem.
Datamat’s software is Web-enabled, which means that it can interface with any Object Database Connectivity-compliant database connected to the Internet. Data mining agents, managed by a Java-based tool called the Mediator, are installed on each computer to be accessed and analyze the database content. This tool generates models based on the data and extracts useful information from each machine. No data components are exchanged among the mining agents. Information fusion is further enhanced through 3-D visualization technology to assist decision-makers in recognizing and understanding the discovered relationships and patterns within the distributed data.
MDA Origins:
Through SBIR Phase I and II contracts, BMDO funded Datamat to develop this technology for use in an advanced warfare management system. Mobile command posts must have accurate, real-time knowledge of friendly, enemy, neutral, and noncombatant activities and locations. This technology is capable of integrating data from different air, land, and sea components, distributing them throughout a global communication grid to communication posts, and presenting them in an optimal way.
Spinoff Applications:
This technology has two primary applications. First, it could make previously unknown observations. Say, for example, you wanted to know what type of car a particular group of travelers preferred to rent when visiting warm destinations. The information needed to answer this question is stored in two different databases located on opposite coasts and operated by two different companies. Using Datamat’s technology, one could access each database, analyze both sets of data, and conclude that this group prefers to rent mid-size white sedans in Florida, but smaller-size green compacts in Georgia. This analysis is done without moving data between locations, thereby ensuring the security of both databases.
Second, this technology could predict future outcomes and events. Say, for example, you wanted to develop a new maintenance schedule for engine repairs for a fleet of 747 jets. The databases with existing maintenance logs are located around the country. Using Datamat’s technology, one could access each database, analyze the data, and find patterns and relationships. With this information, the software could then predict that engines would need complete overhauls every 14 months.
Commercialization:
Datamat is actively marketing this technology, called InferAgent™, to Federal government agencies, state and local government agencies, U.S. armed forces, and Fortune 500 companies. A large-size market opportunity exists, the company says, with law enforcement/domestic security agencies such as the CIA, FBI, and U.S. Customs. Current customers include Lockheed Martin, Northrup Grumman, the Air Force Research Laboratory, and the Army Space Missile & Defense Command.
Commercial markets include fraud detection for health care and financial institutions and security for network operators. With regards to the latter application, Datamat is currently discussing a potential licensing agreement and investment opportunity with a national communications network operator that would use InferAgent technology to detect network intrusions.
Company Profile:
Datamat was founded in 1992 as an information technology company with a focus on software engineering. Since then, the company has grown into a leading high-technology player in the government services market, with many Department of Defense and non-DOD clients, both as a prime contractor and as a subcontractor. Expertise includes software engineering, intelligent systems, enterprise systems, and engineering and training systems. The company headquarters is in McLean, Virginia, with a field office located in Maryland.
In 1999, Datamat spun off the InferX Corporation to market a comprehensive suite of company-developed software tools, including InferAgent, InferView™ (data mining tool with 3-D data visualization capability), and InferXplor™ (an InferView-based tool specifically designed for oil and natural gas companies). Currently, Inferx employs 4 people, but this number is expected to increase to 10 by mid-2002.
Contact Information:
Dr. B.K. Gogia
Datamat Systems Research, Inc.
8260 Greensboro Drive, Suite 120
McLean, VA 22102
Tel: (703) 917-0880
Fax: (703) 917-0563
*UPDATE: Datamat merged with InferX
web: www.inferx.com
|