Scholarlink

<p>Web accessibility evaluation tools are widely used to detect compliance issues with the Web Content Accessibility Guidelines (WCAG), yet their outputs often consist of fragmented rule-level findings that provide limited support for structured analysis and remediation prioritization. This paper introduces WCAG-ODC, a structured framework that adapts Orthogonal Defect Classification (ODC) principles to accessibility defect analysis. The framework transforms raw accessibility findings into structured defect representations using a cause–effect attribute model incorporating defect type, WCAG principle, source element, detection trigger, severity, and impact area. WCAG-ODC was evaluated through a case study of nine public e-services using accessibility findings collected from multiple automated accessibility evaluation tools, with defect-level classification performed on the IBM Equal Access Toolkit dataset. A defect-level dataset of 2,590 accessibility records was constructed and mapped into 31 rule categories and nine higher-level defect families. The results show that Keyboard Access, Interactive Controls, and Structure and Navigation defects dominate the dataset, while severity and impact-area analyses reveal clear relationships between technical defect families and user-facing accessibility barriers. The findings demonstrate that WCAG-ODC enables accessibility defects to be analyzed beyond aggregated tool counts by supporting defect-family analysis, severity profiling, impact mapping, and remediation prioritization. The study concludes that WCAG-ODC provides a systematic and reproducible foundation for accessibility defect classification and accessibility-oriented software quality improvement. Independent coding validation demonstrated almost perfect agreement for Defect Type classification (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\kappa\)</EquationSource> </InlineEquation> = 0.922), supporting the reliability and reproducibility of the proposed framework.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

WCAG-ODC: an orthogonal defect classification framework for web accessibility defects

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月24日

查看全文
  • Mohammed Alannsary

摘要

Web accessibility evaluation tools are widely used to detect compliance issues with the Web Content Accessibility Guidelines (WCAG), yet their outputs often consist of fragmented rule-level findings that provide limited support for structured analysis and remediation prioritization. This paper introduces WCAG-ODC, a structured framework that adapts Orthogonal Defect Classification (ODC) principles to accessibility defect analysis. The framework transforms raw accessibility findings into structured defect representations using a cause–effect attribute model incorporating defect type, WCAG principle, source element, detection trigger, severity, and impact area. WCAG-ODC was evaluated through a case study of nine public e-services using accessibility findings collected from multiple automated accessibility evaluation tools, with defect-level classification performed on the IBM Equal Access Toolkit dataset. A defect-level dataset of 2,590 accessibility records was constructed and mapped into 31 rule categories and nine higher-level defect families. The results show that Keyboard Access, Interactive Controls, and Structure and Navigation defects dominate the dataset, while severity and impact-area analyses reveal clear relationships between technical defect families and user-facing accessibility barriers. The findings demonstrate that WCAG-ODC enables accessibility defects to be analyzed beyond aggregated tool counts by supporting defect-family analysis, severity profiling, impact mapping, and remediation prioritization. The study concludes that WCAG-ODC provides a systematic and reproducible foundation for accessibility defect classification and accessibility-oriented software quality improvement. Independent coding validation demonstrated almost perfect agreement for Defect Type classification ( \(\kappa\)  = 0.922), supporting the reliability and reproducibility of the proposed framework.