Abstract
In response to the escalating threats and vulnerabilities concerning information security, organisations are persistently exploring avenues to enhance their management of information access. Consequently, numerous organisations have embraced biometrics as a solution to achieve heightened authentication security. However, the task of selecting the most suitable biometric attribute for their specific context can prove complex, particularly when combining multiple biometric attributes. The central focus of this study resides in furnishing a comprehensive framework for this complex undertaking, by offering guidance on the thoughtful selection of an suitable, appropriate, and fitting, and robust multimodal biometric authentication system. The intended deployment environment for this system pertains to the public services sector, with a specific emphasis on its implementation within driver licensing centres. This framework is aimed at facilitating well-informed decision-making processes. The deployment of ill-suited technological solutions within an environment, subsequently left unutilised, can entail substantial financial burdens for organisations, and may ultimately culminate in ineffectual system implementations. This study employs a quantitative research design, guided by an interpretive research paradigm to ensure clarity and precision. This encompasses a thorough literature review, supplemented by a requirement survey designed to identify pivotal criteria important for ensuring successful deployment within the pertinent context. A conceptual model, encompassing the identified criteria, is subsequently presented. The research methodology also encompasses the execution of a survey, associated with exploratory factor analysis, which serves as the cornerstone for the assessment of the model. The result of the requirement survey reveals that a noteworthy 71.84% of the respondents advocate for enhancements in the existing person authentication system employed at driver licensing centres. Moreover, they demonstrate a willingness to adopt biometric systems for authentication purposes. To analyse and evaluate the proposed model, this research incorporates a spectrum of complex statistical techniques, including descriptive statistics, factor analysis, chi-square for goodness of fit, Mardia’s coefficient, multivariate kurtosis, and path model analysis. These techniques collectively facilitate a comprehensive organisation and interpretation of the amassed data, fostering an enhanced understanding of the interrelationships between variables, and the overarching alignment of the model. This, in turn, engenders valuable insights that culminate in substantiated conclusions and well-founded recommendations. This study augments the existing body of knowledge by delving into the
iv
confluence of human factors and deployment-related variables that exert influence on public awareness, perception, and the usability of multimodal biometric systems. The formalised framework, termed the Biometric Selection Model Evaluation, proffered within this study, stands to furnish invaluable guidance for public service environments contemplating the deployment of multimodal biometric authentication technologies. This guidance is aimed at mitigating the risk of prospective failed deployments.
Keywords: biometric, multimodal biometric authentication, acceptance, usability, deployment, awareness, security, satisfaction, government, driver licensing centre