Machine learning approach for optimizing usability of healthcare websites

Area of Science:

• Health Informatics
• Computer Science

Background:

• Hospital websites are vital for patient access to medical services and information.
• Assessing website usability is crucial for effective patient engagement in healthcare.
• There is a gap in empirical studies using machine learning for healthcare website usability evaluation.

Purpose of the Study:

• To evaluate the user-friendliness of hospital websites using machine learning models.
• To identify the most effective machine learning algorithms for assessing healthcare website usability.
• To provide insights for optimizing digital healthcare platforms.

Main Methods:

• Utilized Decision Trees, Random Forests, Ridge Regression, and Support Vector Regression models.
• Developed a custom automated tool to assess the usability of 100 hospital websites.
• Employed metrics like R-square, Mean Square Error (MSE), Mean Absolute Error (MAE), and Explained Variance Score (EVS) for evaluation.

Main Results:

• Random Forest Regression achieved 98% accuracy, and Ridge Regression achieved 87% accuracy.
• Key metrics confirmed the predictive accuracy of the machine learning models.
• Cross-validation demonstrated the robustness of Ridge Regression with minimal overfitting.

Conclusions:

• Machine learning, particularly Random Forest and Ridge Regression, is effective for assessing hospital website usability.
• Overall usability is a critical factor in predicting website performance.
• Future research should expand datasets, incorporate user interaction data, and explore deep learning for enhanced assessments.