MTQE.en-he: Machine Translation Quality Estimation for English-Hebrew

Andy Rosenbaum, Assaf Siani, Ilan Kernerman · 2026

Read MTQE.en-he: Machine Translation Quality Estimation for English-Hebrew by Andy Rosenbaum, Assaf Siani, Ilan Kernerman (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-06546.txt.

Description

Judging translation quality for English to Hebrew has had no public benchmark, despite the language pair's practical importance. Andy Rosenbaum and 2 co-authors release MTQE.en-he, the first such benchmark they know of. It pairs hundreds of English segments from a standard translation set with Hebrew machine translations and expert quality scores.

Three human experts rated each translation. The team tested prompting a commercial chatbot, plus two established quality estimation models, and found that combining all three beat the best single model. Full fine-tuning of the models proved unstable, but lightweight tuning methods trained steadily and improved scores. The benchmark and results support future work on this under-resourced pair.

Catalog record

CREATORAndy Rosenbaum, Assaf Siani, Ilan Kernerman
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE22 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-06546

Browse the full library · About the Tower of Babel