Exploring Cross-sentence Contexts for Named Entity Recognition with BERT

TL;DR

This paper investigates the use of cross-sentence context with BERT models to improve named entity recognition performance across multiple languages, introducing a simple voting method to enhance results without modifying BERT architecture.

Contribution

It systematically studies cross-sentence context for NER with BERT and proposes Contextual Majority Voting to boost accuracy across five languages.

Findings

Adding cross-sentence context improves NER performance in all tested languages.

The proposed CMV method further increases accuracy without changing BERT architecture.

Achieved state-of-the-art results on several NER benchmarks.

Abstract

Named entity recognition (NER) is frequently addressed as a sequence classification task where each input consists of one sentence of text. It is nevertheless clear that useful information for the task can often be found outside of the scope of a single-sentence context. Recently proposed self-attention models such as BERT can both efficiently capture long-distance relationships in input as well as represent inputs consisting of several sentences, creating new opportunitites for approaches that incorporate cross-sentence information in natural language processing tasks. In this paper, we present a systematic study exploring the use of cross-sentence information for NER using BERT models in five languages. We find that adding context in the form of additional sentences to BERT input systematically increases NER performance on all of the tested languages and models. Including multiple sentences in each input also allows us to study the predictions of the same sentences in different contexts. We propose a straightforward method, Contextual Majority Voting (CMV), to combine different predictions for sentences and demonstrate this to further increase NER performance with BERT. Our approach does not require any changes to the underlying BERT architecture, rather relying on restructuring examples for training and prediction. Evaluation on established datasets, including the CoNLL'02 and CoNLL'03 NER benchmarks, demonstrates that our proposed approach can improve on the state-of-the-art NER results on English, Dutch, and Finnish, achieves the best reported BERT-based results on German, and is on par with performance reported with other BERT-based approaches in Spanish. We release all methods implemented in this work under open licenses.