Understanding Slang with LLMs: Modelling Cross-Cultural Nuances through Paraphrasing

Wuraola, Ifeoluwa; Dethlefs, Nina; Marciniak, Daniel

Abstract

In the realm of social media discourse, the integration of slang enriches communication, reflecting the sociocultural identities of users. This study investigates the capability of large language models (LLMs) to paraphrase slang within climate-related tweets from Nigeria and the UK, with a focus on identifying emotional nuances. Using DistilRoBERTa as the base-line model, we observe its limited comprehension of slang. To improve cross-cultural understanding , we gauge the effectiveness of leading LLMs: ChatGPT 4, Gemini, and LLaMA3 in slang paraphrasing. While ChatGPT 4 and Gemini demonstrate comparable effectiveness in slang paraphrasing, LLaMA3 shows less coverage , with all LLMs exhibiting limitations in coverage, especially of Nigerian slang. Our findings underscore the necessity for culturally-sensitive LLM development in emotion classification , particularly in non-anglocentric regions.

Citation

Wuraola, I., Dethlefs, N., & Marciniak, D. (2024, November). Understanding Slang with LLMs: Modelling Cross-Cultural Nuances through Paraphrasing. Presented at 2024 Conference on Empirical Methods in Natural Language Processing, Miami, FLorida, USA

• Presentation Conference Type: Conference Paper (published)
• Conference Name: 2024 Conference on Empirical Methods in Natural Language Processing
• Start Date: Nov 12, 2024
• End Date: Nov 16, 2024
• Acceptance Date: Sep 20, 2024
• Online Publication Date: Nov 12, 2024
• Publication Date: Nov 12, 2024
• Deposit Date: Nov 12, 2024
• Publicly Available Date: Jan 14, 2025
• Journal: Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
• Peer Reviewed: Peer Reviewed
• Pages: 15525-15531
• ISBN: 9798891761643
• DOI: https://doi.org/10.18653/v1/2024.emnlp-main.869
• Public URL: https://hull-repository.worktribe.com/output/4912845
• Publisher URL: https://aclanthology.org/2024.emnlp-main.869