Quality indicators in computational linguistics

How assessing generated contents (natural language generation, NLG)? In computational linguistics, two models are available, coming from the subfield of automatic translation: automated metrics and assessments based on human judgments. This website aims to test the quality of generated content using automated tools.

Readability scores

• Readability score (Flesch-Kincaid)
• Levenshtein Distance

Automated metrics

• iBLEU
• Toolkit for machine translation