Machine Learning UI Assessment for Actionable Accessibility Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing user interface (UI) accessibility guidelines lack actionable measurements for visual cues and design elements, making it difficult to create inclusive interfaces for users with varying disabilities, and manual assessment is error-prone and inefficient.
Innovation Solution
A UI assessment system using machine learning techniques to analyze UI elements, generate embeddings, and predict accessibility parameters, providing feedback for UI improvements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual assessment of UI is performed by multiple testers, then accessibility coverage for different disabilities is improved, but time consumption and error rate increase
Solution Approach 1:
The patent replaces manual mechanical assessment processes with an automated machine learning system. The ML model analyzes UI elements, generates embeddings, and predicts accessibility parameters automatically, eliminating the need for multiple human testers while maintaining or improving assessment accuracy and reducing time consumption.
Solution Approach 2:
The patent introduces machine learning models as intermediaries between the UI and accessibility assessment. The ML system acts as a mediator that processes UI representations, generates embeddings, and produces accessibility predictions, bridging the gap between raw UI data and actionable accessibility insights.
2Ease of manufacture
If existing accessibility guidelines are used, then high-level structure recommendations are provided, but actionable measurements for specific design elements are lacking
Solution Approach 1:
The patent segments the UI into individual elements (buttons, text fields, icons, etc.) and analyzes each element separately. The system identifies specific UI components, generates embeddings for each, and provides targeted accessibility measurements for individual elements rather than treating the UI as a monolithic structure.
Solution Approach 2:
The patent transforms UI elements into numerical embeddings and predicts multiple accessibility parameters (error rates, completion rates, time on task, text resizing frequency) for each element. This converts qualitative design guidance into quantitative, measurable parameters that provide actionable feedback.
3Adaptability or versatility
If iterative manual testing is performed to cover varying levels of disability, then comprehensive accessibility coverage is achieved, but productivity and scalability decrease
Solution Approach 1:
The patent creates a universal ML-based assessment system that handles multiple disability types and UI elements simultaneously. The model is trained on diverse accessibility data and can evaluate various UI components (visual, auditory, motor, cognitive aspects) through a single unified process, providing comprehensive coverage without requiring separate testing procedures for each disability type.
Data Source
AI summary
Methods and systems are described herein for a user interface assessment (UI) system that assesses interfaces using machine learning models. The UI assessment system may receive a request for evaluating a UI comprising (1) a representation of the UI and (2) user-defined target values for parameters. The system may identify elements of the UI and structural properties of each element, each structural property controlling presentation of a corresponding element. The system may generate, for each element, (1) a corresponding composite embedding and (2) a corresponding content embedding. The system may generate a graph representation. The system may then generate a graph embedding using an embedding model. The graph embedding may be input into a parameter analysis machine learning model to obtain a corresponding predicted value for each of the parameters and transmit a corresponding predicted value for each of the parameters.