Entity Linking
Entity linking is an important task in natural language processing, aiming to associate entities mentioned in the text (such as famous people, places, or companies) with unique identities in a knowledge base. The goal of this task is to identify and resolve entity mentions in the text and link them to the correct knowledge base entries, thereby achieving accurate understanding and reference of the entities. Entity linking has significant application value in information retrieval, knowledge graph construction, semantic analysis, and other fields, enhancing the system's ability to understand and process text content.
AIDA-CoNLL
SpEL-large (2023)
KILT: AIDA-YAGO2
GENRE
KILT: WNED-CWEB
GENRE
KILT: WNED-WIKI
GENRE
MSNBC
Kannan Ravi et al. (2021)
WiC-TSV
Human
Derczynski
De Cao et al. (2021a)
FUNSD
GeoLayoutLM
EC-FUNSD
N3-Reuters-128
E2E
OKE-2015
E2E
OKE-2016
E2E
KORE50
CoNLL-Aida
Raiman & Raiman 2018
MedMentions
N3-RSS-500
ReLiK-Large
TAC-KBP 2010
AIDA/testc
SpEL-large (2023)
BC7 NLM-Chem
Rare Diseases Mentions in MIMIC-III (Text-to-UMLS)
Rare Diseases Mentions in MIMIC-III
WebQSP-WD
ReFinED
ZESHEL
ArboEL
FIGER
ERNIE
GUM
Rare Diseases Mentions in MIMIC-III Radiology Reports (Text-to-UMLS)
SemEHR+WS (rules+BlueBERT) with tuning number of training data
REBEL