Exploring morphology-aware tokenization: a case study on Spanish language modeling

This paper investigates to what extent the integration of morphological information can improve subword tokenization and thus also language modeling performance. We focus on Spanish, a language with fusional morphology, where subword segmentation can benefit from linguistic structure. Instead of rel...

Descripción completa

Detalles Bibliográficos
Autores: Táboas García, Alba, Przybyła, Piotr, Wanner, Leo
Tipo de recurso: capítulo de libro
Estado:Versión publicada
Fecha de publicación:2025
País:España
Institución:Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya)
Repositorio:Recercat. Dipósit de la Recerca de Catalunya
OAI Identifier:oai:recercat.cat:10230/72720
Acceso en línea:https://hdl.handle.net/10230/72720
http://dx.doi.org/10.18653/v1/2025.emnlp-main.1552
Access Level:acceso abierto
Palabra clave:Morphology-aware tokenization
Spanish language modeling
Descripción
Sumario:This paper investigates to what extent the integration of morphological information can improve subword tokenization and thus also language modeling performance. We focus on Spanish, a language with fusional morphology, where subword segmentation can benefit from linguistic structure. Instead of relying on purely data-driven strategies like Byte Pair Encoding (BPE), we explore a linguistically grounded approach: training a tokenizer on morphologically segmented data. To do so, we develop a semi-supervised segmentation model for Spanish, building gold-standard datasets to guide and evaluate it. We then use this tokenizer to pre-train a masked language model and assess its performance on several downstream tasks. Our results show improvements over a baseline with a standard tokenizer, supporting our hypothesis that morphology-aware tokenization offers a viable and principled alternative for improving language modeling.