Can Combining Demographics and Biometrics Improve De-duplication Performance?

TLDR: The study presents the results when demographic and biometric information are individually processed and complementary information from the two modalities are combined at match score level for de-duplication under different operating scenarios.

Abstract: With the prevalent utilization of citizen databases, an individual has to prove his/her identity for accessing several services such as banking, health care, and social welfare benefits. These databases are now increasingly using demographic and biometric information to uniquely identify the individuals. It protects the core identity of citizens and facilitates them to receive the entitled benefits and rights. It is therefore important that every citizen should enroll only once in the database and be assigned only one unique identifier. De-duplication process prevents an individual from enrolling multiple times in the database. It is essential to understand the importance of constituent information (demographic and biometric) in the de-duplication process. Using a large database, this research attempts to fill the gap in existing literature by analyzing the performance of demographic and biometric information for de-duplication. The study presents the results when demographic and biometric information are individually processed and complementary information from the two modalities are combined at match score level for de-duplication under different operating scenarios.