CAREER: Inclusive Cybersecurity Through the Lens of Accessible Identity and Access Management

点击翻译按钮获取中文摘要

--------------------------------------------------------------------------------------------------------------------------------This Faculty Early Career Development (CAREER) project increases the cybersecurity awareness of individuals identifying as members of racial and ethnic groups historically excluded in Science and Engineering (S&E). Such underrepresentation could lead to reduced opportunities in S&E in which cybersecurity skills are developed. Thus, this research focuses on cybersecurity awareness concerning the appropriate use and management of identity credentials for user authentication. This focus is important as credential-related account compromises are among the most common types of cyberattacks, while members of historically excluded racial and ethnic groups are more frequently targeted by cyberattacks. Further, while research on inclusive authentication is growing, inclusive user authentication schemes for historically excluded racial and ethnic groups are understudied. Through integrated research and education activities, this project develops novel, inclusive user authentication systems to reduce cybersecurity risk, with race and ethnicity central foci. Outcomes of this project also provide cybersecurity-focused education material for K-12 and university students and the broader population, thereby expanding the nation's cybersecurity defense.This project engages individuals from historically excluded racial and ethnic groups in S&E to inform the design, implementation, and appropriate use of inclusive authentication systems. The research team is conducting a series of focus groups to identify current use and perceptions of existing authentication systems, how various cybersecurity practices are applied in everyday life, and how various factors, including racial equity and cultural values, shape perceptions of computing and cybersecurity. Performance of authentication systems, including knowledge-based and AI-derived biometric systems, is also investigated to identify data-driven and algorithmic biases. This performance analysis is facilitated through a multi-session data collection, through which mock credentials and physical and behavioral biometric cues are gathered from volunteering research participants. With these insights and data, novel authentication systems that address poor generalization, failure trends, and authentication errors associated with subgroups of users, the authentication type, or the authentication model are developed using a variety of approaches, including soft biometric classification, feature selection, and multimodal fusion. This CAREER project informs the design and implementation of authentication systems that contribute to inclusive and accessible cybersecurity solutions, transforms state-of-the-art authentication systems by exposing when and how they isolate certain groups and identifying biases in knowledge and biometric-based authentication systems, contributes a novel, diverse dataset for cybersecurity research, and informs future directions for inclusive identity and access management.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.