paper-with-me

Papers

ROMEVA: Geometry-Preserving Vocabulary Expansion for Roman Urdu Language Models

2026-06-21 · Mahnoor Khan, Afsheen Asif, Milhan Afzal Khan, Seemab Latif, Mehwish Fatima arxiv

Multilingual Language Models like mBERT are widely used for low-resource NLP, yet their adaptation to morphologically inconsistent languages such as Roman Urdu remains underexplored. Roman Urdu spelling variation causes severe sub-word fragmentation, averaging 1.50 sub-words per token. We propose \textit{ROMEVA} (Roman Urdu Embedding-preserving Vocabulary Adaptation), which combines sub-word-average initialization and a PCA-guided anchor loss to stabilize embeddings during vocabulary expansion. Using a 36,130-comment Roman Urdu corpus, we add 500 highly fragmented tokens to mBERT and compare naive fine-tuning, sub-word-aware fine-tuning, and \textit{ROMEVA}. While \textit{ROMEVA} most effectively preserves the pretrained embedding space, naive fine-tuning achieves the strongest downstream sentiment classification performance. These findings reveal a disconnect between embedding stability and downstream performance, suggesting that stronger adaptation may be preferable to strict embedding preservation in morphologically inconsistent languages.

📄 PDF Abstract BibTeX arXiv:2606.22478

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Evaluating the Wordnet and CoRoLa-based Word Embedding Vectors for Romanian as Resources in the Task of Microworlds Lexicon Expansion

2019-07-01 · GWC 2019 7 · Elena Irimia, Maria Mitrofan, Verginica Mititelu

Within a larger frame of facilitating human-robot interaction, we present here the creation of a core vocabulary to be learned by a robot. It is extracted from two tokenised and lemmatized scenarios pertaining to two ima…

On Romanization for Model Transfer Between Scripts in Neural Machine Translation

2020-09-30 · Findings of the Association for Computational Linguistics 2020 · Chantal Amrhein, Rico Sennrich

Transfer learning is a popular strategy to improve the quality of low-resource machine translation. For an optimal transfer of the embedding layer, the child and parent model should share a substantial part of the vocabu…

Machine TranslationTransfer LearningTranslation

IITP-MT at WAT2021: Indic-English Multilingual Neural Machine Translation using Romanized Vocabulary

2021-08-01 · ACL (WAT) 2021 8 · Ramakrishna Appicharla, Kamal Kumar Gupta, Asif Ekbal, Pushpak Bhattacharyya

This paper describes the systems submitted to WAT 2021 MultiIndicMT shared task by IITP-MT team. We submit two multilingual Neural Machine Translation (NMT) systems (Indic-to-English and English-to-Indic). We romanize al…

Machine TranslationNMTTranslation

Romanization-based Large-scale Adaptation of Multilingual Language Models

2023-04-18 · Sukannya Purkayastha, Sebastian Ruder, Jonas Pfeiffer, Iryna Gurevych 외

Large multilingual pretrained language models (mPLMs) have become the de facto state of the art for cross-lingual transfer in NLP. However, their large-scale deployment to many languages, besides pretraining data scarcit…

Cross-Lingual TransferTransliteration

Romanization Encoding For Multilingual ASR

2024-07-05 · Wen Ding, Fei Jia, Hainan Xu, Yu Xi 외

We introduce romanization encoding for script-heavy languages to optimize multilingual and code-switching Automatic Speech Recognition (ASR) systems. By adopting romanization encoding alongside a balanced concatenated to…

Automatic Speech RecognitionAutomatic Speech Recognition (ASR)Language ModelingLanguage Modelling+2