Neural Architectures for Nested NER through Linearization

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... (read more)

PDF Abstract ACL 2019 PDF ACL 2019 Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK USES EXTRA
TRAINING DATA
RESULT BENCHMARK
Nested Named Entity Recognition ACE 2004 seq2seq+BERT+Flair F1 84.40 # 3
Nested Mention Recognition ACE 2004 seq2seq+BERT+Flair F1 84.40 # 3
Named Entity Recognition ACE 2004 seq2seq+BERT+Flair F1 84.40 # 3
Nested Named Entity Recognition ACE 2005 seq2seq+BERT+Flair F1 84.33 # 2
Named Entity Recognition ACE 2005 seq2seq+BERT+Flair F1 84.33 # 4
Nested Mention Recognition ACE 2005 seq2seq+BERT+Flair F1 84.33 # 2
Named Entity Recognition CoNLL 2002 (Dutch) Straková et al., 2019 F1 92.7 # 5
Named Entity Recognition CoNLL 2002 (Spanish) Straková et al., 2019 F1 88.8 # 4
Named Entity Recognition CoNLL 2003 (English) LSTM-CRF+ELMo+BERT+Flair F1 93.38 # 8
Named Entity Recognition CoNLL 2003 (German) Straková et al., 2019 F1 85.1 # 5
Named Entity Recognition GENIA seq2seq+BERT+Flair F1 78.31 # 2
Nested Named Entity Recognition GENIA seq2seq+BERT+Flair F1 78.31 # 3

Methods used in the Paper


METHOD TYPE
Sigmoid Activation
Activation Functions
Tanh Activation
Activation Functions
LSTM
Recurrent Neural Networks
BiLSTM
Bidirectional Recurrent Neural Networks
ELMo
Word Embeddings
Residual Connection
Skip Connections
Attention Dropout
Regularization
Linear Warmup With Linear Decay
Learning Rate Schedules
Weight Decay
Regularization
GELU
Activation Functions
Dense Connections
Feedforward Networks
Adam
Stochastic Optimization
WordPiece
Subword Segmentation
Softmax
Output Functions
Dropout
Regularization
Multi-Head Attention
Attention Modules
Layer Normalization
Normalization
Scaled Dot-Product Attention
Attention Mechanisms
BERT
Language Models