Communication Dans Un Congrès Année : 2025

Common Ground, Diverse Roots: The Difficulty of Classifying Common Examples in Spanish Varieties

Résumé

Variations in languages across geographic regions or cultures are crucial to address to avoid biases in NLP systems designed for culturally sensitive tasks, such as hate speech detection or dialog with conversational agents. In languages such as Spanish, where varieties can significantly overlap, many examples can be valid across them, which we refer to as common examples. Ignoring these examples may cause misclassifications, reducing model accuracy and fairness. Therefore, accounting for these common examples is essential to improve the robustness and representativeness of NLP systems trained on such data. In this work, we address this problem in the context of Spanish varieties. We use training dynamics to automatically detect common examples or errors in existing Spanish datasets. We demonstrate the efficacy of using predicted label confidence for our Datamaps (Swayamdipta et al., 2020) implementation for the identification of hardto-classify examples, especially common examples, enhancing model performance in variety identification tasks. Additionally, we introduce a Cuban Spanish Variety Identification dataset with common examples annotations developed to facilitate more accurate detection of Cuban and Caribbean Spanish varieties. To our knowledge, this is the first dataset focused on identifying the Cuban, or any other Caribbean, Spanish variety.

Fichier principal
Vignette du fichier
Spanish_variants_vardial-3.pdf (761.18 Ko) Télécharger le fichier

Dates et versions

hal-04868010 , version 1 (06-01-2025)

Licence

Identifiants

  • HAL Id : hal-04868010 , version 1

Citer

Javier Alejandro Lopetegui Gonzalez, Arij Riabi, Djamé Seddah. Common Ground, Diverse Roots: The Difficulty of Classifying Common Examples in Spanish Varieties. VarDial 2025 - Twelfth Workshop on NLP for Similar Languages, Varieties and Dialects co-located with COLING 2025, Jan 2025, Abu Dhabi, United Arab Emirates. ⟨hal-04868010⟩
157 Consultations
204 Téléchargements

Partager

  • More