Entity Disambiguation On Aida Conll

Metrics

In-KB Accuracy

Results

Performance results of various models on this benchmark

		Paper Title
confidence-order	95.0	Global Entity Disambiguation with BERT
DCA-SL + Triples	94.94	Evaluating the Impact of Knowledge Graph Context on Entity Disambiguation Models
DeepType	94.88	DeepType: Multilingual Entity Linking by Neural Type System Evolution
NTEE	94.7	Learning Distributed Representations of Texts and Entities from Knowledge Base
DCA-SL (2019)(et al., [2019c])	94.64	Learning Dynamic Context Augmentation for Global Entity Linking
Fang et al. (2019) (et al., [2019e])	94.3	Joint Entity Linking with Deep Reinforcement Learning
This work+CtxLSTMs+LDC+MPCM	94.0	Neural Cross-Lingual Entity Linking
ReFinED	93.9	ReFinED: An Efficient Zero-shot-capable Approach to End-to-End Entity Linking
Chen et al. (2020) (et al, 2020)	93.54	Improving Entity Linking by Modeling Latent Entity Type Information
GENRE	93.3	Autoregressive Entity Retrieval
Wikipedia2Vec-GBRT	93.1	Joint Learning of the Embedding of Words and Entities for Named Entity Disambiguation
ELDEN	93.0	ELDEN: Improved Entity Linking Using Densified Knowledge Graphs
NER4EL	92.5	Named Entity Recognition for Entity Linking: What Works and What’s Next
Global	92.22	Deep Joint Entity Disambiguation with Local Neural Attention
Wikipedia2Vec	91.5	Joint Learning of the Embedding of Words and Entities for Named Entity Disambiguation
KBED	90.4	Improving Entity Disambiguation by Reasoning over a Knowledge Base
Le& Titov (2019) (Le and Titov, 2019)	89.66	Boosting Entity Linking Performance by Leveraging Unlabeled Documents
Hoffart et al.	82.29	Robust Disambiguation of Named Entities in Text
BERT-Entity-Sim (local & global) AIDA-B	-	Improving Entity Linking by Modeling Latent Entity Type Information
Bootleg	-	Bootleg: Chasing the Tail with Self-Supervised Named Entity Disambiguation

0 of 20 row(s) selected.

Command Palette

Entity Disambiguation On Aida Conll

Metrics

Results