As more of our lives are measured—from the traces we leave in our online activities to the biometric observations captured via increasingly pervasive imaging systems—it becomes possible to draw associations between aspects of identity to produce a fuller picture of an individual’s identity and to predict characteristics that are not expressed in direct measurements. Moreover, our exposure to cyber risks will continue to grow as elements of our cyber-identities are spread widely across online environments. Other elements of identity that exist in the natural world, such as biometrics, may enable connections across the cyber/natural world divide through their association with cyber identities. Once any transition across the cyber/physical divide has been made, the amount of additional insight about an identity that can be inferred is significant. Fusing cyber and physical identities can open new opportunities and risks in cyber defense, privacy protection, and law enforcement.&
A multinational team, of which the Pacific Northwest National Laboratory is a part, is constructing a new model for contemporary identity called Superidentity. Superidentity is the fusion of physical and digital identities that connects cyber, biological, biographical, and psychological attributes—it is the core identity of an individual to which all of their personas and physical traits belong. Through SuperIdentity, can we predict physical appearance from an online avatar, gait from touchscreen biometrics, or psychological characteristics from an online dating profile? Our objective in this research is to work towards a rich computational model for identity that encompasses aspects that we reveal both in the real world and in the cyber world. The result will be a more complete and dynamic picture of who someone is, enabling greater input into identification and attribution determinations as well as to identity protection activities. The SuperIdentity model also allows each piece of the identity puzzle to be assessed for its reliability, enabling a certainty index to be associated with the identity decision. This talk will review the SuperIdentity model and associated visualization system for exploring the inferences that can be drawn across elements of identity, describe the use cases that supported its development, and share initial results in identify inference capabilities.
Pizza lunch available at 11:45 am, RMCIC 1301
ripple [atsymbol] cert.org (Kimberly Ripple)