Turkish Delights: a Dataset on Turkish Euphemisms

TL;DR

This paper introduces the Turkish PETs dataset—the first of its kind for Turkish euphemism detection—by compiling 122 Turkish PETs, collecting contextual examples, and annotating them with four-way labels. It demonstrates a binary classification approach using transformer models, comparing multilingual (XLM-RoBERTa, mBERT) and Turkish-specific (BERTurk, ELECTRA Turkish) architectures, with macro-F1 as a primary metric. Key findings show that Turkish-specific models generally outperform multilingual ones, and ELECTRA Turkish yields the best results, while a balanced dataset improves detection performance. The work provides a foundation for Turkish euphemism analysis and suggests future directions in explainability, larger datasets, and cross-lingual transfer to broaden euphemism detection in low-resource languages.

Abstract

Euphemisms are a form of figurative language relatively understudied in natural language processing. This research extends the current computational work on potentially euphemistic terms (PETs) to Turkish. We introduce the Turkish PET dataset, the first available of its kind in the field. By creating a list of euphemisms in Turkish, collecting example contexts, and annotating them, we provide both euphemistic and non-euphemistic examples of PETs in Turkish. We describe the dataset and methodologies, and also experiment with transformer-based models on Turkish euphemism detection by using our dataset for binary classification. We compare performances across models using F1, accuracy, and precision as evaluation metrics.