The Impact of Data Corruption on Named Entity Recognition for Low-resourced Languages
This addresses data challenges in NLP for low-resourced languages, providing insights for efficient model training, though it is incremental as it builds on existing methods.
The study systematically measured the impact of data quantity and quality on pre-trained language models for low-resourced languages, finding that fewer completely-labeled sentences outperform more sentences with missing labels, and models achieve strong performance with only 10% of training data across ten languages and four models.
Data availability and quality are major challenges in natural language processing for low-resourced languages. In particular, there is significantly less data available than for higher-resourced languages. This data is also often of low quality, rife with errors, invalid text or incorrect annotations. Many prior works focus on dealing with these problems, either by generating synthetic data, or filtering out low-quality parts of datasets. We instead investigate these factors more deeply, by systematically measuring the effect of data quantity and quality on the performance of pre-trained language models in a low-resourced setting. Our results show that having fewer completely-labelled sentences is significantly better than having more sentences with missing labels; and that models can perform remarkably well with only 10% of the training data. Importantly, these results are consistent across ten low-resource languages, English, and four pre-trained models.