Putting Evaluation in Context: Contextual Embeddings Improve Machine Translation Evaluation
Year: 2019
Type: conference-paper
Abstract: Accurate, automatic evaluation of machine translation is critical for system tuning, and evaluating progress in the field.We proposed a simple unsupervised metric, and additional supervised metrics which rely on contextual word embeddings to encode the translation and reference sentences.We find that these models rival or surpass all existing metrics in the WMT 2017 sentence-level and systemlevel tracks, and our trained model has a substantially higher correlation with human judgements than all existing metrics on the WMT 2017 to-English sentence level dataset.
DOI: 10.18653/v1/p19-1269
Authors: Nitika Mathur, Timothy Baldwin, Trevor Cohn
Institution: The University of Melbourne
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W2953287808
