Generating unlabelled data for a tri-training approach in a low resourced NER task - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Generating unlabelled data for a tri-training approach in a low resourced NER task

Hugo Boulanger
Thomas Lavergne
Sophie Rosset

Résumé

Training a tagger for Named Entity Recognition (NER) requires a substantial amount of labeled data in the task domain. Manual labeling is a tedious and complicated task. Semisupervised learning methods can reduce the quantity of labeled data necessary to train a model. However, these methods require large quantities of unlabeled data, which remains an issue in many cases. We address this problem by generating unlabeled data. Large language models have proven to be powerful tools for text generation. We use their generative capacity to produce new sentences and variations of the sentences of our available data. This generation method, combined with a semi-supervised method, is evaluated on CoNLL and I2B2. We prepare both of these corpora to simulate a low resource setting. We obtain significant improvements for semisupervised learning with synthetic data against supervised learning on natural data.
Fichier principal
Vignette du fichier
2022.deeplo-1.4.pdf (227.95 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-03813272 , version 1 (13-10-2022)

Identifiants

Citer

Hugo Boulanger, Thomas Lavergne, Sophie Rosset. Generating unlabelled data for a tri-training approach in a low resourced NER task. Third Workshop on Deep Learning for Low-Resource Natural Language Processing, Jul 2022, Hybrid, Seattle, United States. pp.30-37, ⟨10.18653/v1/2022.deeplo-1.4⟩. ⟨hal-03813272⟩
141 Consultations
44 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More