HiTZ at VarDial 2025 NorSID: Overcoming Data Scarcity with Language Transfer and Automatic Data Annotation

2024-12-13

AI Summary

To address performance bottlenecks in low-resource Norwegian dialectal NLU tasks—specifically intent classification, slot filling, and dialect identification—this work proposes a joint language transfer and domain adaptation framework. Building upon XLM-R, it leverages multilingual xSID data for cross-lingual transfer and performs domain-adaptive fine-tuning specifically on dialect identification. This study presents the first systematic evaluation of cross-lingual pretraining’s generalization capability and domain specificity for under-resourced dialectal tasks. Through multi-task fine-tuning, data distribution consistency analysis, and ablation studies, the model achieves a 2.3% improvement in F1 score for intent classification and slot filling on the NorSID test set over baseline models, while attaining 89.7% accuracy on dialect identification—ranking among the top three on the official VarDial 2025 NorSID leaderboard.

Technology Category

• Natural Language Processing: Machine Translation, Multilinguality, Cross-Lingual NLP
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Application Domains: Humanities & Computational Social Science

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

In this paper we present our submission for the NorSID Shared Task as part of the 2025 VarDial Workshop (Scherrer et al., 2025), consisting of three tasks: Intent Detection, Slot Filling and Dialect Identification, evaluated using data in different dialects of the Norwegian language. For Intent Detection and Slot Filling, we have fine-tuned a multitask model in a cross-lingual setting, to leverage the xSID dataset available in 17 languages. In the case of Dialect Identification, our final submission consists of a model fine-tuned on the provided development set, which has obtained the highest scores within our experiments. Our final results on the test set show that our models do not drop in performance compared to the development set, likely due to the domain-specificity of the dataset and the similar distribution of both subsets. Finally, we also report an in-depth analysis of the provided datasets and their artifacts, as well as other sets of experiments that have been carried out but did not yield the best results. Additionally, we present an analysis on the reasons why some methods have been more successful than others; mainly the impact of the combination of languages and domain-specificity of the training data on the results.

Problem

• Intention Recognition
• Slot Filling
• Dialect Identification

Innovation

• Multi-task Learning Model
• Cross-lingual Generalization
• Automatic Data Annotation