BLEU Neighbors: A Reference-less Approach to Automatic Evaluation
This addresses the bottleneck of expensive human evaluation for open-ended NLG tasks, enabling faster iteration, though it is incremental as it builds on existing BLEU metrics.
The paper tackles the problem of evaluating natural language generation models without references by proposing BLEU Neighbors, a nearest neighbors model using BLEU as a kernel function, which achieves lower mean squared error and higher Spearman correlation with ground truth than individual human annotators on datasets for chitchat dialogue and open-ended sentence generation.
Evaluation is a bottleneck in the development of natural language generation (NLG) models. Automatic metrics such as BLEU rely on references, but for tasks such as open-ended generation, there are no references to draw upon. Although language diversity can be estimated using statistical measures such as perplexity, measuring language quality requires human evaluation. However, because human evaluation at scale is slow and expensive, it is used sparingly; it cannot be used to rapidly iterate on NLG models, in the way BLEU is used for machine translation. To this end, we propose BLEU Neighbors, a nearest neighbors model for estimating language quality by using the BLEU score as a kernel function. On existing datasets for chitchat dialogue and open-ended sentence generation, we find that -- on average -- the quality estimation from a BLEU Neighbors model has a lower mean squared error and higher Spearman correlation with the ground truth than individual human annotators. Despite its simplicity, BLEU Neighbors even outperforms state-of-the-art models on automatically grading essays, including models that have access to a gold-standard reference essay.