# Feature-Dependent Confusion Matrices for Low-Resource NER Labeling with Noisy Labels

arXiv: https://arxiv.org/abs/1910.06061 · HTML: https://mlsanity.com/paper/1910.06061/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 14 Oct 2019.

## Built on by (most cited first, 5 of 5)

- [Is BERT Robust to Label Noise? A Study on Learning with Noisy Labels in Text Classification](https://mlsanity.com/paper/2204.09371/) (2022)
- [Analysing the Noise Model Error for Realistic Noisy Label Data](https://mlsanity.com/paper/2101.09763/) (2021)
- [ANEA: Distant Supervision for Low-Resource Named Entity Recognition](https://mlsanity.com/paper/2102.13129/) (2021)
- [ULF: Unsupervised Labeling Function Correction using Cross-Validation for Weak Supervision](https://mlsanity.com/paper/2204.06863/) (2022)
- [Low Resource Pipeline for Spoken Language Understanding via Weak Supervision](https://mlsanity.com/paper/2206.10559/) (2022)

## Abstract (arXiv)

In low-resource settings, the performance of supervised labeling models can be improved with automatically annotated or distantly supervised data, which is cheap to create but often noisy. Previous works have shown that significant improvements can be reached by injecting information about the confusion between clean and noisy labels in this additional training data into the classifier training. However, for noise estimation, these approaches either do not take the input features (in our case word embeddings) into account, or they need to learn the noise modeling from scratch which can be difficult in a low-resource setting. We propose to cluster the training data using the input features and then compute different confusion matrices for each cluster. To the best of our knowledge, our approach is the first to leverage feature-dependent noise modeling with pre-initialized confusion matrices. We evaluate on low-resource named entity recognition settings in several languages, showing that our methods improve upon other confusion-matrix based methods by up to 9%.