This interface offers you to test the quality of automatically generated texts through quality indicators.

Each of the proposed metrics is explained and documented. Some required to have two texts (source text, written by a human, target text, written by a software) in order to make comparisons.

This project is developed in the context of a PhD dissertation about "Quality and uses of natural language generation in the field of journalism".

Université Libre de Bruxelles (ULB), Sciences et technologies de l'information et de la communication STIC 2016