Exploring the Use of AI for Enhanced Accessibility Testing of Web Solutions

Area of Science:

• Computer Science
• Human-Computer Interaction
• Web Accessibility

Background:

• Digital products and services often present accessibility barriers for people with disabilities, limiting their participation in the information society.
• Mainstream web solutions require improvements to ensure equal access for all users.
• Artificial Intelligence (AI) offers potential solutions for enhancing digital accessibility and user experience.

Purpose of the Study:

• To explore AI applications for creating accessible digital solutions.
• To investigate how AI techniques can streamline web accessibility testing.
• To demonstrate the effectiveness of AI in enhancing conformance testing against accessibility guidelines.

Main Methods:

• Selected four success criteria from Web Content Accessibility Guidelines (WCAG) that necessitate significant manual testing.
• Developed four prototypes utilizing open-source machine-learning models.
• Focused on automating and enhancing accessibility conformance testing.

Main Results:

• AI-based techniques show potential to significantly reduce manual effort in web accessibility testing.
• Prototypes demonstrated the feasibility of using AI for enhancing conformance testing.
• The research indicates a pathway toward more efficient accessibility evaluation.

Conclusions:

• AI holds significant promise for advancing web accessibility and user experience.
• AI-driven automation can make accessibility testing more efficient and effective.
• Further optimization and evaluation of AI prototypes are needed to fully realize their potential in ensuring digital inclusion.