Metraipolis › archive › dc_7c81ce2a31fb
Hate Speech Classification In Roman Urdu: A Comparative Study On Parameter Efficient Fine-Tuning And Prompt Engineering — arXiv:2608.21408
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.21408
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T07:38:38.885Z
- SHA-256 of the archived text
19160f3e6ab560d11794b880d9f22f513e9d4ffe2aa9959f18dd1fe2387e10e7- Stable URL of this page
https://metraipolis.com/documento/dc_7c81ce2a31fb
How to cite
Hate Speech Classification In Roman Urdu: A Comparative Study On Parameter Efficient Fine-Tuning And Prompt Engineering — arXiv:2608.21408 — arXiv (en), https://arxiv.org/abs/2608.21408 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7c81ce2a31fb (Metraipolis archive dc_7c81ce2a31fb, sha256 19160f3e6ab560d11794b880d9f22f513e9d4ffe2aa9959f18dd1fe2387e10e7)
BibTeX
@misc{dc7c81ce2a31fb,
title = {Hate Speech Classification In Roman Urdu: A Comparative Study On Parameter Efficient Fine-Tuning And Prompt Engineering — arXiv:2608.21408},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.21408},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7c81ce2a31fb; sha256 19160f3e6ab560d11794b880d9f22f513e9d4ffe2aa9959f18dd1fe2387e10e7},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.21408. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Due to the widespread accessibility of the internet and social media, toxic and hateful con-tent has grown exponentially, causing significant distress and negative societal impacts. Ro-man Urdu, a low-resource language used in Pakistan and among Urdu-speaking communities worldwide, presents additional challenges because of its informal grammar, inconsistent sen-tence structures, and multiple variations in word spellings. This research aims to identify the most effective techniques for hate speech classification in such low-resource settings with limited data. To address this, the study investigates and compares the latest approaches, in-cluding prompt tuning, parameter-efficient fine-tuning (PEFT) using LoRA, and prompt en-gineering, under various experimental configurations. To achieve this objective, four exper-iments were designed. The first experiment involved direct inferencing with LLMs without any fine-tuning, to evaluate how well these models understand Roman Urdu in a zero-shot setting, especially given limited data. The second experiment utilized parameter-efficient fine-tuning (PEFT) with LoRA, which updates only a small subset of parameters, thereby reducing computational cost. The third experiment explored prompt tuning with both mixed and manually crafted prompts, using very small sets of training examples relative to the entire dataset, making it computationally efficient as well. Finally, the fourth experiment applied prompt engineering through zero-shot and few-shot learning, relying solely on care-fully designed instruction prompts for classification without further training. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-25 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.21408 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-18T07:37:52.636Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).