Abstract:There is an increasing global emphasis on information accessibility, with numerous researchers actively developing automated tools to detect and repair accessibility issues, thereby ensuring that individuals with diverse abilities can independently access software products and services. However, current research still encounters significant challenges in two key areas: the absence of a comprehensive taxonomy of accessibility issue types, and the lack of comprehensive analysis of the capabilities of detection and repair tools, as well as the status of corresponding datasets. To address these challenges, this paper introduces the Accessibility Issue Analysis (AIA) framework. Utilizing this framework, we develop a comprehensive taxonomy that categorizes 55 types of accessibility issues across four pivotal dimensions: Perceivability, Operability, Understandability, and Robustness. This taxonomy has been rigorously recognized through a questionnaire survey (n=130). Building on this taxonomy, we conduct an in-depth analysis of existing detection and repair tools, as well as the status of corresponding datasets. In terms of tools, our findings indicate that 14 detection tools can identify 31 issue types, achieving a 56.3% rate (31/55). Meanwhile, 9 repair tools address just 13 issue types, with a 23.6% rate. In terms of datasets, those for detection tools cover 21 issue types, at a 38.1% coverage rate, whereas those for repair tools cover only 7 types, at a 12.7% coverage rate.

What problem does this paper attempt to address?

The main problems that this paper attempts to solve include: 1. **Lack of a comprehensive classification system for types of accessibility problems**: - Current research has not yet systematically collected and classified the accessibility problems that users encounter when using mobile applications and websites. A comprehensive classification system will help to understand these problems more in - depth and assist developers in designing more accessible software products. 2. **Insufficient comprehensive analysis of the capabilities of detection and repair tools**: - Although a variety of automated detection and repair tools have been developed, the capabilities of these tools have not been thoroughly evaluated. Specifically, existing research lacks a comprehensive analysis of these tools in terms of identifying and repairing different types of accessibility problems. 3. **Unclear status of relevant data sets**: - The data sets used in the research have a limited coverage and cannot comprehensively reflect all types of accessibility problems. High - quality data sets are crucial for tool performance, so a comprehensive evaluation of existing data sets is required. To solve the above problems, the paper proposes the "Accessibility Issues Analysis (AIA)" framework and has carried out the following work: - **Constructed a comprehensive classification system for types of accessibility problems**: Through literature review and questionnaire surveys, 55 types of accessibility problems were determined and classified according to the WCAG 2.1 standard. - **Analyzed the capabilities of existing detection and repair tools**: Evaluated the identification and repair capabilities of 14 detection tools and 9 repair tools for different types of accessibility problems. - **Evaluated the status of existing data sets**: Analyzed the data sets used for detection and repair tools and found that their coverage was limited, being 38.1% and 12.7% respectively. By answering three key research questions (RQs), this paper aims to enhance the understanding of the classification of accessibility problems, reveal the limitations of existing tools, and provide valuable references for future research and practice.

SoK: Detection and Repair of Accessibility Issues