Relation Extraction On Docred

评估指标

Ign F1

评测结果

各个模型在此基准测试上的表现结果

模型名称	F1	Ign F1	Paper Title	Repository
MIUK	59.99	58.05	Multi-view Inference for Relation Extraction with Uncertain Knowledge
EncAttAgg	58.7	56.7	Improving Document-level Relation Extraction via Contextualizing Mention Representations and Weighting Mention Pairs
SAIS-BERT-base	62.77	60.96	SAIS: Supervising and Augmenting Intermediate Steps for Document-Level Relation Extraction
SSAN-RoBERTa-large+Adaptation	65.92	63.78	Entity Structure Within and Throughout: Modeling Mention Dependencies for Document-Level Relation Extraction
ATLOP + LogiRE	61.45	59.48	Learning Logic Rules for Document-level Relation Extraction
MRN+BERT	61.74	59.52	-	-
BERT-base	53.22	56.17	Fine-tune Bert for DocRED with Two-step Process
GAIN+SIEF	62.29	59.87	Document-Level Relation Extraction with Sentences Importance Estimation and Focusing
GLRE-XLNet-Large	59.0	56.8	Global-to-Local Neural Networks for Document-Level Relation Extraction
SSAN-RoBERTa-base	59.94	57.71	Entity Structure Within and Throughout: Modeling Mention Dependencies for Document-Level Relation Extraction
DRN-BERT-base	61.37	59.15	Discriminative Reasoning for Document-level Relation Extraction
ATLOP + KIRE	61.39	59.35	Enhancing Document-level Relation Extraction by Entity Knowledge Injection
MRN	58.46	56.19	-	-
ATLOP-BERT-base	61.30	59.31	Document-Level Relation Extraction with Adaptive Thresholding and Localized Context Pooling
CFER-GloVe	55.75	53.43	Coarse-to-Fine Entity Representations for Document-level Relation Extraction
DocRED-CNN	42.33	36.44	DocRED: A Large-Scale Document-Level Relation Extraction Dataset
Two-Step+BERT-base	53.92	54.42	Fine-tune Bert for DocRED with Two-step Process
CFER-BERT-base	59.82	57.89	Coarse-to-Fine Entity Representations for Document-level Relation Extraction
LSR+GloVe	54.18	52.15	Reasoning with Latent Structure Refinement for Document-Level Relation Extraction
DocuNet-RoBERTa-large	64.55	62.4	Document-level Relation Extraction as Semantic Segmentation

0 of 62 row(s) selected.