Automatic Quality Estimation for Natural Language Generation: Ranting (Jointly Rating and Ranking)
This work addresses the problem of evaluating NLG outputs automatically for researchers and practitioners, offering incremental improvements over existing methods.
The paper tackles automatic quality estimation for natural language generation by developing a recurrent neural network system that jointly learns to assign ratings and provide pairwise rankings, using synthetic data to improve performance. This approach achieves a 12% increase in correlation with human ratings over the previous benchmark and a 4% accuracy increase on a ranking dataset.
We present a recurrent neural network based system for automatic quality estimation of natural language generation (NLG) outputs, which jointly learns to assign numerical ratings to individual outputs and to provide pairwise rankings of two different outputs. The latter is trained using pairwise hinge loss over scores from two copies of the rating network. We use learning to rank and synthetic data to improve the quality of ratings assigned by our system: we synthesise training pairs of distorted system outputs and train the system to rank the less distorted one higher. This leads to a 12% increase in correlation with human ratings over the previous benchmark. We also establish the state of the art on the dataset of relative rankings from the E2E NLG Challenge (Dušek et al., 2019), where synthetic data lead to a 4% accuracy increase over the base model.