Date of Award

5-7-2026

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Computer Science

First Advisor

Philip Huff

Abstract

Screen reader users face significant barriers when using web-based cybersecurity tools, however, the accessibility of these interfaces has received minimal systematic research attention. Existing automated evaluation tools assess Web Content Accessibility Guidelines (WCAG) conformance but do not measure the practical operability of complex and domain-specific interfaces for users of assistive technology. This dissertation presents the Deciphering Interfaces for Your Accessibility (DIYA) framework, an open-source automated auditing framework that evaluates cybersecurity tool web interfaces across eight normalized dimensions: WCAG conformance, Accessible Rich Internet Applications (ARIA) usage, semantic structure, keyboard operability, form accessibility, dynamic content accessibility, interaction cost, and screen reader readiness. DIYA combines these dimensions using empirically informed composite weights and integrates the results with the National Initiative for Cybersecurity Education (NICE) Workforce Framework to map accessibility barriers to workforce role impacts. The framework was applied to four open-source cybersecurity tools: Arkime (network traffic analysis), Wazuh (intrusion detection), OpenVAS (vulnerability scanning), and the Malware Information Sharing Platform (MISP, threat intelligence sharing). Across 102 audited pages, DIYA identified 7,393 accessibility issues, of which 55.2\% were detected exclusively by DIYA's non-WCAG auditors and were invisible to axe-core. Keyboard reachability ranged from 0.393 (MISP) to 0.936 (Wazuh), focus traps appeared on five of the nine Arkime pages, and semantic landmarks were systematically absent across three of the four tools. The NICE Workforce Framework coverage of Protect and Defend competency items ranged from 3.1\% (Arkime) to 92.2\% (Wazuh), demonstrating that workforce inclusion for screen reader users depends on the tools organizations deploy. A Monte Carlo sensitivity analysis confirmed that composite rankings remained stable under weight perturbation (median Kendall's $\tau$ = 0.52--0.89) and that keyboard operability was the most influential dimension. A structured validation walkthrough on eight representative pages confirmed 90.6\% concordance with DIYA's original findings. These findings establish that WCAG conformance testing alone substantially underestimates accessibility barriers in cybersecurity tool interfaces and that accessible cybersecurity tool design is achievable when prioritized during development.

Download

Share

COinS