AI Accessibility Remediation for Complex Web UI Elements
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing web accessibility assessment tools struggle to identify complex web elements and require manual effort due to their lack of defined structure or semantics, leading to increased development time and cost, and existing automated tools can only verify a limited number of WCAG checkpoints.
Innovation Solution
A processor-implemented method and system that automatically identify applicable accessibility guidelines, non-compliances, and remediation techniques by analyzing user interface elements, simulating user interactions, and applying fixes to the Document Object Model (DOM) in real-time, utilizing artificial intelligence and machine learning to handle complex web elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If traditional HTML DOM-based identification methods are used, then basic web elements can be identified easily, but complex web elements (carousels, menu bars, date pickers, sliders, expand/collapsible panels, tree views) cannot be properly identified and assessed
Solution Approach 1:
The patent introduces an intermediary component (machine learning model/AI system) that acts as a mediator between the HTML DOM structure and the accessibility assessment process. This intermediary translates complex UI element structures into meaningful accessibility assessments by learning patterns from training data, enabling proper identification of elements like carousels, menu bars, and date pickers that lack standard HTML semantics.
2Measurement precision
If manual effort is used to identify complex web elements, then accessibility assessment accuracy improves, but development time and cost increase
Solution Approach 1:
The system implements self-service by automatically identifying and assessing complex web elements through AI/ML-based detection. The machine learning model autonomously analyzes UI element structures, determines their types, and performs accessibility assessments without requiring manual identification by developers or accessibility specialists, thereby maintaining high accuracy while reducing time and cost.
Solution Approach 2:
The patent replaces the mechanical manual process of identifying and assessing complex web elements with an automated AI/ML-based system. This substitution transitions from human-dependent manual inspection to an automated intelligent system that can process and assess complex UI elements efficiently, significantly reducing development time while maintaining or improving assessment accuracy.
3Extent of automation
If existing automated accessibility tools are used, then testing phase automation is achieved, but only 25-29% of WCAG checkpoints can be verified
Solution Approach 1:
The patent fundamentally changes the operational parameters of accessibility assessment by transitioning from static HTML DOM-based analysis to dynamic AI/ML-based UI element analysis. This parameter change enables the system to handle complex web elements and achieve comprehensive WCAG checkpoint verification coverage (96% as claimed) by adapting to the actual rendered UI structure rather than relying solely on source code semantics.
Solution Approach 2:
The system employs a composite approach combining multiple methodologies: traditional HTML DOM analysis, AI/ML-based UI element identification, and automated remediation techniques. This composite system integrates the strengths of different approaches to achieve both high automation extent and comprehensive WCAG checkpoint verification, overcoming the limitations of any single method.
4Ease of operation
If developers manually implement accessibility, then complete control over remediation is maintained, but development complexity and time increase significantly
Solution Approach 1:
The system provides self-service by automatically identifying accessibility non-compliances and generating remediation code. The AI/ML model not only detects issues but also autonomously produces fix code, reducing developer burden and simplifying the accessibility implementation process while maintaining the ability to review and approve automated recommendations.
Solution Approach 2:
The patent implements preliminary action by automatically generating remediation code during the development process itself, rather than requiring separate manual intervention later. The system proactively identifies accessibility issues and provides fix code in real-time, allowing developers to address accessibility concerns as part of the normal development workflow, thereby reducing overall development complexity and time.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
Accessibility in software engineering is treated as expensive, time consuming and hence adoption of accessibility, is a challenge despite stringent timelines and regulatory requirements published around the world. Moreover, cost of implementing accessibility increases project cost due to manual intervention and dependency on niche skills, which is scarce in industry. Embodiments of the present disclosure provide system and method for automated identification of applicable accessibility guidelines and determination of remediation techniques for fixing issues in webpages, wherein webpages are rendered, and applicable accessibility guidelines are identified based on user interface elements (UI) comprised in the webpages. Further, content associated with rendered webpages are analyzed using the applicable accessibility applicable guidelines to identify webpage non-compliance issue(s) and remediation technique(s) thereof. Fixes for the non-compliance issues are determined based on the webpages and applied on a document object model (DOM) based on a current state associated with the webpage and/or UI elements.