We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: To Ship or Not to Ship: An Extensive Evaluation of Automatic Metrics for Machine Translation

Abstract: Automatic metrics are commonly used as the exclusive tool for declaring the superiority of one machine translation system's quality over another. The community choice of automatic metric guides research directions and industrial developments by deciding which models are deemed better. Evaluating metrics correlations with sets of human judgements has been limited by the size of these sets. In this paper, we corroborate how reliable metrics are in contrast to human judgements on -- to the best of our knowledge -- the largest collection of judgements reported in the literature. Arguably, pairwise rankings of two systems are the most common evaluation tasks in research or deployment scenarios. Taking human judgement as a gold standard, we investigate which metrics have the highest accuracy in predicting translation quality rankings for such system pairs. Furthermore, we evaluate the performance of various metrics across different language pairs and domains. Lastly, we show that the sole use of BLEU impeded the development of improved models leading to bad deployment decisions. We release the collection of 2.3M sentence-level human judgements for 4380 systems for further analysis and replication of our work.
Comments: Accepted to WMT 2021 research papers
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2107.10821 [cs.CL]
  (or arXiv:2107.10821v2 [cs.CL] for this version)

Submission history

From: Tom Kocmi [view email]
[v1] Thu, 22 Jul 2021 17:22:22 GMT (2050kb,D)
[v2] Mon, 13 Sep 2021 23:31:16 GMT (2047kb,D)

Link back to: arXiv, form interface, contact.