Massively Scalable Themes, Entities and Relationships (MASTER)

Award Information
Agency:
Department of Defense
Branch
Defense Advanced Research Projects Agency
Amount:
$749,999.00
Award Year:
2010
Program:
SBIR
Phase:
Phase II
Contract:
W31P4Q-11-C-0041
Agency Tracking Number:
08SB2-0495
Solicitation Year:
2008
Solicitation Topic Code:
SB082-026
Solicitation Number:
2008.2
Small Business Information
DECISIVE ANALYTICS Corporation
1235 South Clark Street, Suite 400, Arlington, VA, -
Hubzone Owned:
N
Socially and Economically Disadvantaged:
N
Woman Owned:
N
Duns:
036593457
Principal Investigator:
Jonathon Day
Senior Engineer
(703) 414-5015
jonathon.day@dac.us
Business Contact:
Kelly McClelland
VP, Administration
(703) 414-5024
kelly.mcclelland@dac.us
Research Institution:
n/a
Abstract
Discovering the important evidence in the form of activities, actors and relationships in a sea of open source data requires the ability to extract and correlate seemingly unrelated pieces of data, distinguish that data from the noise of harmless civilian activity and find the hidden attributes and relationships that characterize the true threat. To meet these requirements, the DAC BOBCAT Team proposes to extend a new suite of algorithms developed that enable current NLP applications to be immediately available across all levels of military intelligence. We call this approach the Massively Scalable Themes, Entities, and Relationships (MASTER). In the MASTER approach, we overcome the scalability limitations of current NLP approaches while also enabling the tactical warfighter to focus queries based on discovered context and relations. The development of the MASTER approach for the tactical warfighter will result in a suite of algorithms that will support all levels of the fight. The current suite of BOBCAT algorithms, which already advance the state of the art in statistical theme discovery, executes on enterprise platforms at the strategic and operational levels. The MASTER algorithms will specifically focus on the tactical levels with limited computing footprints and large amounts of open-source data.

* information listed above is at the time of submission.

Agency Micro-sites

US Flag An Official Website of the United States Government