all AI news
Lost in Translationese? Reducing Translation Effect Using Abstract Meaning Representation. (arXiv:2304.11501v2 [cs.CL] UPDATED)
cs.CL updates on arXiv.org arxiv.org
Translated texts bear several hallmarks distinct from texts originating in
the language. Though individual translated texts are often fluent and preserve
meaning, at a large scale, translated texts have statistical tendencies which
distinguish them from text originally written in the language
("translationese") and can affect model performance. We frame the novel task of
translationese reduction and hypothesize that Abstract Meaning Representation
(AMR), a graph-based semantic representation which abstracts away from the
surface form, can be used as an interlingua to …
abstract arxiv cs.cl language lost meaning performance representation scale statistical text them translated translation