To improve interpretability, the SHAP framework was used being a activity-theory–based technique that assigns each feature a measurable effect on pred