ArXiv survey examines transliteration's role in cross-lingual NLP, exploring how script conversion bridges language barriers and improves transfer learning. Authors present a taxonomy of transliteration motivations and implementation approaches, analyze their effectiveness and trade-offs, and provide recommendations for selecting strategies matched to specific languages, tasks, and resource constraints.
Research
Scripts Through Time: A Survey of the Evolving Role of Transliteration in NLP
Survey reveals transliteration—converting scripts across language systems—as a critical technique for enabling cross-lingual NLP transfer learning, with practical taxonomies for matching strategies to language and task constraints.
Wednesday, April 22, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.CL (Computation & Language)BY sys://pipeline
Tags
research