SubRegWeigh: Effective and Efficient Annotation Weighing with Subword Regularization

TL;DR

SubRegWeigh introduces a fast, subword regularization-based method for annotation error detection and weighting, significantly reducing computation time while improving performance in NLP tasks like document classification and NER.

Contribution

It presents a novel, efficient annotation weighing method using subword regularization that outperforms existing approaches in speed and accuracy.

Findings

Performs annotation weighting 4-5 times faster than previous methods.

Improves accuracy in document classification and named entity recognition.

Effectively detects pseudo-incorrect labels as annotation errors.

Abstract

NLP datasets may still contain annotation errors, even when they are manually annotated. Researchers have attempted to develop methods to automatically reduce the adverse effect of errors in datasets. However, existing methods are time-consuming because they require many trained models to detect errors. This paper proposes a time-saving method that utilizes a tokenization technique called subword regularization to simulate multiple error detection models for detecting errors. Our proposed method, SubRegWeigh, can perform annotation weighting four to five times faster than the existing method. Additionally, SubRegWeigh improved performance in document classification and named entity recognition tasks. In experiments with pseudo-incorrect labels, SubRegWeigh clearly identifies pseudo-incorrect labels as annotation errors. Our code is available at https://github.com/4ldk/SubRegWeigh .