Human Assessment: Arguably, the most reliable method to determine the quality of a simplification consists of asking human judges to rate it. It is common practice to evaluate a model’s output on three criteria: grammaticality, meaning preservation, and simplicity.

Automatic Metrics: Some of these metrics are based on comparing the automatic simplifications to manually produced references; others compute the readability of the text based on psycholinguistic metrics; whereas others are trained on specially annotated data so as to learn to predict the quality or usefulness of the simplification being evaluated.