ACCESS: Automated Comparison and Clustering of Entity Signatures

21st Century Technologies’ (21CT) ACCESS (Automated Comparison and Clustering of Entity SignatureS) research effort addresses the issue of comparing entities such as human personas and networks, so that a more complete assessment of at-risk entities can be made within and across the various domains in which those entities interact. The Phase I effort of ACCESS will provide an effective similarity measure and clustering algorithm for the automated comparison of human signatures that arise within and across modalities and mediums so that personas of interest can be discovered among large databases of gathered intelligence. This similarity measure will be used to create higher fidelity personas and provide the means to detect asymmetric actors/at-risk entities that manifest behavior types of interest. Key components of the ACCESS Phase I effort include: 1) Identification of one or more useful similarity measures for clustering personas, 2) Experimental data to inform assessments of the effectiveness of the identified measures, and 3) Preliminary results identifying at-risk signatures. This builds directly upon existing 21CT technologies that generate signatures from raw data, providing the customer with the ability to determine signature similarity for the identification of persons of interest.