SeqScore: Addressing Barriers to Reproducible Named Entity Recognition Evaluation

TL;DR

This paper introduces SeqScore, a software tool and guidelines to improve the reproducibility of named entity recognition evaluation by promoting transparency and addressing scoring discrepancies.

Contribution

It presents simple guidelines and a software package, SeqScore, to enhance reproducibility and address issues causing evaluation inconsistencies in NER.

Findings

Unreported scoring differences significantly affect NER results.

SeqScore mitigates many causes of replication failures.

Transparency in scoring procedures is crucial for reproducibility.

Abstract

To address a looming crisis of unreproducible evaluation for named entity recognition, we propose guidelines and introduce SeqScore, a software package to improve reproducibility. The guidelines we propose are extremely simple and center around transparency regarding how chunks are encoded and scored. We demonstrate that despite the apparent simplicity of NER evaluation, unreported differences in the scoring procedure can result in changes to scores that are both of noticeable magnitude and statistically significant. We describe SeqScore, which addresses many of the issues that cause replication failures.