关系提取
关系抽取是预测句子中实体属性及其相互关系的任务，旨在识别和分类文本中的实体间关系。该任务对于构建关系知识图谱至关重要，能够显著提升结构化搜索、情感分析、问答系统和文本摘要等自然语言处理应用的效果。
DocRED
DREEAM
TACRED
DeepStruct multi-task w/ finetune
SemEval-2010 Task-8
ACE 2005
Multi-turn QA
CoNLL04
REBEL
Adverse Drug Events (ADE) Corpus
ITER
WebNLG
PFN
ChemProt
SciBert (Finetune)
NYT11-HRL
RERE
ACE 2004
PL-Marker
CDR
SAISORE+CR+ET-SciBERT
NYT10-HRL
ReRe
FUNSD
GDA
Re-TACRED
SpanBERT
NYT
NYT Corpus
KGPOOL
ReDocRED
DREEAM
SemEval 2018 Task 10
SVM with GloVe
NYT21
DDI
KeBioLM
DWIE
GAD
BioLinkBERT (large)
NYT-single
ETL-Span
NYT29
SciERC
PFN
SKE
ReRe (exact)
TACRED-Revisited
BioRED
PubMedBERT
Dataset: Relationship extraction for knowledge graph creation from biomedical literature (Gene-Disease relationships)
FewRel
ERNIE
NYT24
WDec
REBEL
SemEval-2010 Task 8
LLM-QA4RE (XXLarge)
WLPC
SpanRel
2010 i2b2/VA
Spark NLP
2012 i2b2 Temporal Relations
Spark NLP
2018 n2c2 posology
Spark NLP
ADE Corpus
PFN
Dataset: Relationship extraction for knowledge graph creation from biomedical literature (Gene-Disease relationships) n
DuIE
Google RE
JNLPBA
SciBERT (SciVocab)
LPSC-contains
LPSC-hasproperty
Stacked_LinkedBERT
MUC6
iDepNN
PGR
Spark NLP
sciERC-sent
RELA
Wikipedia-Wikidata relations
ContextAtt
WNUT 2020
Baseline