TY - GEN
T1 - A Framework to Identify People in Unstructured Environments Incorporating Biometrics
AU - Mason, Janelle
AU - Chatterjee, Prosenjit
AU - Roy, Kaushik
AU - Esterline, Albert
N1 - Publisher Copyright:
© 2019, Springer Nature Switzerland AG.
PY - 2019
Y1 - 2019
N2 - We outline our computational framework for identity. We have a prototype web application, but this paper is a conceptual level. The interest is in identity as an equivalence relation and how information can be evidence for identity hypotheses. Our account is based on the situation theory of Barwise and Perry. We consider a (legal) identity case to be a constellation of situations, and we indicate how the structure of such a case facilitates discounting and combining evidence using Dempster-Shafer theory. Semantic Web resources are used to capture the structure of evidence as it relates to situations. We have developed OWL ontologies and use the concepts therein defined in RDF triple stores to capture case data. URIs (as used in the Semantic Web) are used for unambiguous references to individuals. We sketch a scenario that uses two biometric modalities in an uncontrolled environment and show how our framework applies. Recently, biometrics has gained the limelight as a means to identify individuals, but much else may be available for this task, including sensor data, witness reports, and data on file. To our knowledge, this is the only framework that in principle can accommodate any kind of evidence for identity. It is not an alternative to biometrics, but rather provides a way to incorporate biometrics into a larger context.
AB - We outline our computational framework for identity. We have a prototype web application, but this paper is a conceptual level. The interest is in identity as an equivalence relation and how information can be evidence for identity hypotheses. Our account is based on the situation theory of Barwise and Perry. We consider a (legal) identity case to be a constellation of situations, and we indicate how the structure of such a case facilitates discounting and combining evidence using Dempster-Shafer theory. Semantic Web resources are used to capture the structure of evidence as it relates to situations. We have developed OWL ontologies and use the concepts therein defined in RDF triple stores to capture case data. URIs (as used in the Semantic Web) are used for unambiguous references to individuals. We sketch a scenario that uses two biometric modalities in an uncontrolled environment and show how our framework applies. Recently, biometrics has gained the limelight as a means to identify individuals, but much else may be available for this task, including sensor data, witness reports, and data on file. To our knowledge, this is the only framework that in principle can accommodate any kind of evidence for identity. It is not an alternative to biometrics, but rather provides a way to incorporate biometrics into a larger context.
KW - Argumentation schemes
KW - Biometrics
KW - Dempster-Shafer theory
KW - Evidence
KW - Identity
KW - Semantic Web
UR - https://www.scopus.com/pages/publications/85069861820
U2 - 10.1007/978-3-030-24900-7_5
DO - 10.1007/978-3-030-24900-7_5
M3 - Conference contribution
SN - 9783030248994
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 65
EP - 75
BT - Security, Privacy, and Anonymity in Computation, Communication, and Storage - SpaCCS 2019 International Workshops, Proceedings
A2 - Wang, Guojun
A2 - Feng, Jun
A2 - Bhuiyan, Md Zakirul Alam
A2 - Lu, Rongxing
PB - Springer Verlag
T2 - 12th International Conference on Security, Privacy, and Anonymity in Computation, Communication, and Storage, SpaCCS 2019
Y2 - 14 July 2019 through 17 July 2019
ER -