Human Factors in Large-Scale Biometric Systems: A Study of the Human Factors Related to Errors in Semiautomatic Fingerprint Biometrics

Abstract

The purpose of this paper is to demonstrate the importance of considering human factors in large-scale, complex biometric systems. A team of 19 board-certified latent print examiners conducted 1620 latent fingerprint image formatting tasks, 1797 encoding tasks, and 146 388 side-by-side comparison tasks of latent prints with potential matching tenprint candidates. Examiner feedback from ten encoding mistakes and 13 comparison mistakes provided significant data to demonstrate the deserved consideration of relating the Department of Defense (DoD) Human Factors Classification System (HFACS) to semiautomatic fingerprint biometrics. The increase in match rate of 10% and 7%, respectively, for encoding and comparison when verifications are conducted on these tasks provides striking evidence of the risks involved if large-scale biometric system integrators or owners design and operate biometric systems based solely on single human examiner conclusions without ample consideration of the error recovery mechanism of second-examiner involvement for low-quality data.