CULTURAL APPROPRIATENESS OF MACHINE-GENERATED CONGRATULATIONS IN UZBEK AND RUSSIAN: A CORPUS-INFORMED EVALUATION FRAMEWORK

Main Article Content

Yuldasheva Sabina Yusufovna

Abstract

 Large language models are increasingly used to write congratulatory messages, yet their output is usually judged by fluency rather than by conformity to the etiquette norms of a particular linguaculture. This paper proposes a corpus-informed framework for evaluating the cultural appropriateness of machine-generated congratulations in Uzbek and Russian. Norm profiles are derived from a reference corpus of authentic congratulatory texts [N texts] and describe, for each communicative situation, the expected form of address, level of formality, obligatory rhetorical moves, and typical set of wish concepts. Generated texts are compared with these profiles along six criteria, combining automatic measures with expert ratings. The framework is tested on [N] texts produced by [models] in response to prompts varying the occasion, the age and status of the addressee, and the channel. The study shows [main findings]. The framework contributes to the evaluation of culturally sensitive text generation for low-resource languages such as Uzbek.

Article Details

Section

Articles

References

1.parrots: Can language models be too big? // Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency. — 2021. — P. 610–623.

2.Brown P., Levinson S. C. Politeness: Some Universals in Language Usage. — Cambridge: Cambridge University Press, 1987.

3.Formanovskaya N. I. Rechevoy etiket i kul’tura obshcheniya [Speech etiquette and the culture of communication]. — Moscow: Vysshaya shkola, 1989.

4.Hakimov M. O‘zbek pragmalingvistikasi asoslari [Fundamentals of Uzbek pragmalinguistics]. — Tashkent: Akademnashr, 2013.

5.Hershcovich D. et al. Challenges and strategies in cross-cultural NLP // Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics. — 2022. — P. 6997–7013.

6.Joshi P., Santy S., Budhiraja A., Bali K., Choudhury M. The state and fate of linguistic diversity and inclusion in the NLP world // Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. — 2020. — P. 6282–6293.

7.Krippendorff K. Content Analysis: An Introduction to Its Methodology. — 2nd ed. — Thousand Oaks: Sage, 2004.

8.Searle J. R. A classification of illocutionary acts // Language in Society. — 1976. — Vol. 5, No. 1. — P. 1–23.

9.Spencer-Oatey H. (ed.) Culturally Speaking: Culture, Communication and Politeness Theory. — 2nd ed. — London: Continuum, 2008.

10.Wierzbicka A. Cross-Cultural Pragmatics: The Semantics of Human Interaction. — Berlin; New York: Mouton de Gruyter, 1991.