Relation Extraction On Ace 2005

评估指标

Cross Sentence

Relation classification 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	Cross Sentence	Relation classification F1	Paper Title	Repository
Dual Pointer Network(multi-head)	No	80.8	Dual Pointer Network for Fast Extraction of Multiple Relations in a Sentence	-
Multi-turn QA	No	-	Entity-Relation Extraction as Multi-Turn Question Answering
RNN+CNN	No	67.7	Combining Neural Networks and Log-linear Models to Improve Relation Extraction	-
SPTree	No	-	End-to-End Relation Extraction using LSTMs on Sequences and Tree Structures
TablERT	No	-	Named Entity Recognition and Relation Extraction using Enhanced Table Filling by Contextualized Representations
MRC4ERE++	No	-	Asking Effective and Diverse Questions: A Machine Reading Comprehension based Framework for Joint Entity-Relation Extraction
MGE	No	-	A Multi-Gate Encoder for Joint Entity and Relation Extraction	-
HySPA	No	-	HySPA: Hybrid Span Generation for Scalable Text-to-Graph Extraction
ASP+T5-3B	Yes	-	Autoregressive Structured Prediction with Language Models
Walk-based model	No	64.2	A Walk-based Model on Entity Graphs for Relation Extraction
DYGIE++	Yes	-	Entity, Relation, and Event Extraction with Contextualized Span Representations
Table-Sequence	No	-	Two are Better than One: Joint Entity and Relation Extraction with Table-Sequence Encoders
PL-Marker	Yes	-	Packed Levitated Marker for Entity and Relation Extraction
CNN	No	61.3	-	-
Dual Pointer Network	No	80.5	Relation Extraction among Multiple Entities Using a Dual Pointer Network with a Multi-Head Attention Mechanism	-
GoLLIE	-	-	GoLLIE: Annotation Guidelines improve Zero-Shot Information-Extraction
MRT	No	-	Extracting Entities and Relations with Joint Minimum Risk Training	-
Joint w/ Global	No	-	-	-
Span-level	No	-	Span-Level Model for Relation Extraction	-
Hierarchical Multi-task	No	-	A Hierarchical Multi-task Approach for Learning Embeddings from Semantic Tasks

0 of 30 row(s) selected.