Neural Architectures for Nested NER through Linearization
Jana Straková, Milan Straka, Jan Hajič
Code Available — Be the first to reproduce this paper.
ReproduceCode
Abstract
We propose two neural network architectures for nested named entity recognition (NER), a setting in which named entities may overlap and also be labeled with more than one label. We encode the nested labels using a linearized scheme. In our first proposed approach, the nested labels are modeled as multilabels corresponding to the Cartesian product of the nested labels in a standard LSTM-CRF architecture. In the second one, the nested NER is viewed as a sequence-to-sequence problem, in which the input sequence consists of the tokens and output sequence of the labels, using hard attention on the word whose label is being predicted. The proposed methods outperform the nested NER state of the art on four corpora: ACE-2004, ACE-2005, GENIA and Czech CNEC. We also enrich our architectures with the recently published contextual embeddings: ELMo, BERT and Flair, reaching further improvements for the four nested entity corpora. In addition, we report flat NER state-of-the-art results for CoNLL-2002 Dutch and Spanish and for CoNLL-2003 English.
Tasks
Benchmark Results
| Dataset | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| ACE 2004 | seq2seq+BERT+Flair | F1 | 84.4 | — | Unverified |
| ACE 2005 | seq2seq+BERT+Flair | F1 | 84.33 | — | Unverified |
| CoNLL 2002 (Dutch) | Straková et al., 2019 | F1 | 92.7 | — | Unverified |
| CoNLL 2002 (Spanish) | Straková et al., 2019 | F1 | 88.8 | — | Unverified |
| CoNLL 2003 (English) | LSTM-CRF+ELMo+BERT+Flair | F1 | 93.38 | — | Unverified |
| CoNLL 2003 (German) | Straková et al., 2019 | F1 | 85.1 | — | Unverified |
| GENIA | seq2seq+BERT+Flair | F1 | 78.31 | — | Unverified |