Modality Matching Matters: Calibrating Language Distances for Cross-Lingual Transfer in URIEL+
Identifikátory výsledku
Kód výsledku v IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F26%3AJCICN3XT" target="_blank" >RIV/00216208:11320/26:JCICN3XT - isvavai.cz</a>
Výsledek na webu
<a href="http://arxiv.org/abs/2510.19217" target="_blank" >http://arxiv.org/abs/2510.19217</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.48550/arXiv.2510.19217" target="_blank" >10.48550/arXiv.2510.19217</a>
Alternativní jazyky
Jazyk výsledku
angličtina
Název v původním jazyce
Modality Matching Matters: Calibrating Language Distances for Cross-Lingual Transfer in URIEL+
Popis výsledku v původním jazyce
Existing linguistic knowledge bases such as URIEL+ provide valuable geographic, genetic and typological distances for cross-lingual transfer but suffer from two key limitations. One, their one-size-fits-all vector representations are ill-suited to the diverse structures of linguistic data, and two, they lack a principled method for aggregating these signals into a single, comprehensive score. In this paper, we address these gaps by introducing a framework for type-matched language distances. We propose novel, structure-aware representations for each distance type: speaker-weighted distributions for geography, hyperbolic embeddings for genealogy, and a latent variables model for typology. We unify these signals into a robust, task-agnostic composite distance. In selecting transfer languages, our representations and composite distances consistently improve performance across a wide range of NLP tasks, providing a more principled and effective toolkit for multilingual research.
Název v anglickém jazyce
Modality Matching Matters: Calibrating Language Distances for Cross-Lingual Transfer in URIEL+
Popis výsledku anglicky
Existing linguistic knowledge bases such as URIEL+ provide valuable geographic, genetic and typological distances for cross-lingual transfer but suffer from two key limitations. One, their one-size-fits-all vector representations are ill-suited to the diverse structures of linguistic data, and two, they lack a principled method for aggregating these signals into a single, comprehensive score. In this paper, we address these gaps by introducing a framework for type-matched language distances. We propose novel, structure-aware representations for each distance type: speaker-weighted distributions for geography, hyperbolic embeddings for genealogy, and a latent variables model for typology. We unify these signals into a robust, task-agnostic composite distance. In selecting transfer languages, our representations and composite distances consistently improve performance across a wide range of NLP tasks, providing a more principled and effective toolkit for multilingual research.
Klasifikace
Druh
O - Ostatní výsledky
CEP obor
—
OECD FORD obor
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Návaznosti výsledku
Projekt
—
Návaznosti
—
Ostatní
Rok uplatnění
2025
Kód důvěrnosti údajů
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů