Fact · Languages and accents
The 2002 paper that introduced BLEU proposed a quick, inexpensive, language-independent automatic method that correlates highly with human evaluation.
Checked against its source on October 17, 2026
In the source's own words
“that is quick, inexpensive, and language-independent, that correlates highly with human”
BLEU: a Method for Automatic Evaluation of Machine Translation, Kishore Papineni et al., Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, 2002, Abstract.
Open the original- What it measures
- BLEU is the score used for most speech translation benchmarks. It is an automatic measure, so researchers can score many systems without paying human judges for each run.
- What it does not show
- The paper validates BLEU against human judgement for text translation in 2002. It is not a claim about how a phone app feels to a user.