%0 Generic %1 schweter2020flert %A Schweter, Stefan %A Akbik, Alan %D 2020 %K flair model nlp %T FLERT: Document-Level Features for Named Entity Recognition %U http://arxiv.org/abs/2011.06993 %X Current state-of-the-art approaches for named entity recognition (NER) using BERT-style transformers typically use one of two different approaches: (1) The first fine-tunes the transformer itself on the NER task and adds only a simple linear layer for word-level predictions. (2) The second uses the transformer only to provide features to a standard LSTM-CRF sequence labeling architecture and thus performs no fine-tuning. In this paper, we perform a comparative analysis of both approaches in a variety of settings currently considered in the literature. In particular, we evaluate how well they work when document-level features are leveraged. Our evaluation on the classic CoNLL benchmark datasets for 4 languages shows that document-level features significantly improve NER quality and that fine-tuning generally outperforms the feature-based approaches. We present recommendations for parameters as well as several new state-of-the-art numbers. Our approach is integrated into the Flair framework to facilitate reproduction of our experiments.

@misc{schweter2020flert, abstract = {Current state-of-the-art approaches for named entity recognition (NER) using BERT-style transformers typically use one of two different approaches: (1) The first fine-tunes the transformer itself on the NER task and adds only a simple linear layer for word-level predictions. (2) The second uses the transformer only to provide features to a standard LSTM-CRF sequence labeling architecture and thus performs no fine-tuning. In this paper, we perform a comparative analysis of both approaches in a variety of settings currently considered in the literature. In particular, we evaluate how well they work when document-level features are leveraged. Our evaluation on the classic CoNLL benchmark datasets for 4 languages shows that document-level features significantly improve NER quality and that fine-tuning generally outperforms the feature-based approaches. We present recommendations for parameters as well as several new state-of-the-art numbers. Our approach is integrated into the Flair framework to facilitate reproduction of our experiments.}, added-at = {2020-12-09T09:38:28.000+0100}, author = {Schweter, Stefan and Akbik, Alan}, biburl = {https://www.bibsonomy.org/bibtex/20eac7f9a26da78373e0bdb56a985420f/parismic}, interhash = {8aa02b08252d00c55f928f22769b6bd9}, intrahash = {0eac7f9a26da78373e0bdb56a985420f}, keywords = {flair model nlp}, note = {cite arxiv:2011.06993}, timestamp = {2020-12-09T09:38:28.000+0100}, title = {FLERT: Document-Level Features for Named Entity Recognition}, url = {http://arxiv.org/abs/2011.06993}, year = 2020 }