English

Machine Translation Evaluation with BERT Regressor

Computation and Language 2019-07-31 v1

Abstract

We introduce the metric using BERT (Bidirectional Encoder Representations from Transformers) (Devlin et al., 2019) for automatic machine translation evaluation. The experimental results of the WMT-2017 Metrics Shared Task dataset show that our metric achieves state-of-the-art performance in segment-level metrics task for all to-English language pairs.

Keywords

Cite

@article{arxiv.1907.12679,
  title  = {Machine Translation Evaluation with BERT Regressor},
  author = {Hiroki Shimanaka and Tomoyuki Kajiwara and Mamoru Komachi},
  journal= {arXiv preprint arXiv:1907.12679},
  year   = {2019}
}

Comments

6 pages

R2 v1 2026-06-23T10:34:17.989Z