Summary of Hitz at Vardial 2025 Norsid: Overcoming Data Scarcity with Language Transfer and Automatic Data Annotation, by Jaione Bengoetxea et al.
HiTZ at VarDial 2025 NorSID: Overcoming Data Scarcity with Language Transfer and Automatic Data Annotation
by Jaione Bengoetxea, Mikel Zubillaga, Ekhi Azurmendi, Maite Heredia, Julen Etxaniz, Markel Ferro, Jeremy Barnes
First submitted to arxiv on: 13 Dec 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents our submission to the NorSID Shared Task at the 2025 VarDial Workshop, focusing on three tasks: Intent Detection, Slot Filling, and Dialect Identification. We fine-tuned a multitask model in a cross-lingual setting using xSID data available in 17 languages for Intent Detection and Slot Filling. For Dialect Identification, we used a model fine-tuned on the development set to achieve the highest scores. Our results show that our models perform similarly well on the test set as they do on the development set, likely due to domain-specificity and dataset similarity. We also analyze the provided datasets and experiments, highlighting the impact of language combination and training data on results. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about a competition where researchers try to make computers better at understanding different types of language. The goal is to improve how well computers can detect what someone means (Intent Detection), fill in missing information (Slot Filling), and identify which type of Norwegian dialect someone is speaking (Dialect Identification). To do this, the team used special computer models that can learn from lots of languages at once. They also looked at why some methods worked better than others. |
Keywords
» Artificial intelligence » Intent detection
