Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study investigates the effectiveness of two script unification approaches—uroman romanizer and Common Turkic Script (CTS)—for enhancing cross-lingual transfer in Turkic languages with different scripts. By training matched fastText models on transliterated Wikipedia data from 11 Turkic languages, the authors evaluate performance on named entity recognition (NER) and part-of-speech (POS) tagging tasks. The findings reveal that while both methods outperform monolingual baselines, their effectiveness varies significantly based on language-specific factors and the level of supervision available for the target language.
Script unification strategies can dramatically enhance cross-lingual performance, but their effectiveness hinges on language-specific characteristics and supervision levels.
Closely related languages written in different scripts expose little surface overlap to multilingual models, limiting cross-lingual transfer. We compare two approaches to script unification: the general-purpose uroman romanizer and the family-specific Common Turkic Script (CTS). We train matched fastText models on transliterated Wikipedia corpora from 11 Turkic languages and evaluate them on WikiANN named entity recognition and Universal Dependencies part-of-speech tagging. CTS and uroman show no significant difference on NER, while both substantially outperform the official monolingual fastText baselines. POS results reveal no universal winner: language-specific differences are associated with the cross-lingual character n-gram coverage induced by each representation, while within-language coverage becomes more important when target-language supervision is available. Although CANINE-c achieves higher overall POS averages, the substantially simpler fastText-based systems remain competitive on several treebanks. Overall, the effectiveness of script unification depends on the language, the induced subword overlap, and the available supervision.