LLMs systematically reduce politeness, intimacy, and formality variation in translation, and a retrieval-augmented prompting method that supplies native style exemplars improves style alignment without hurting content quality on GPT-4.
TyDiP: A Dataset for Politeness Classification in Nine Typologically Diverse Languages
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
We study politeness phenomena in nine typologically diverse languages. Politeness is an important facet of communication and is sometimes argued to be cultural-specific, yet existing computational linguistic study is limited to English. We create TyDiP, a dataset containing three-way politeness annotations for 500 examples in each language, totaling 4.5K examples. We evaluate how well multilingual models can identify politeness levels -- they show a fairly robust zero-shot transfer ability, yet fall short of estimated human accuracy significantly. We further study mapping the English politeness strategy lexicon into nine languages via automatic translation and lexicon induction, analyzing whether each strategy's impact stays consistent across languages. Lastly, we empirically study the complicated relationship between formality and politeness through transfer experiments. We hope our dataset will support various research questions and applications, from evaluating multilingual models to constructing polite multilingual agents.
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Towards Style Alignment in Cross-Cultural Translation
LLMs systematically reduce politeness, intimacy, and formality variation in translation, and a retrieval-augmented prompting method that supplies native style exemplars improves style alignment without hurting content quality on GPT-4.