The “CockrACE” corpus consists of 140 news articles annotated with mentions of entities and their coreference links, as well as relation mentions for the evaluation of relation extraction (RE) experiments. Three semantic relations have been annotated, each of them dealing with people's family relationships (marriages, brother/sister, parent/child).
nlp
natural-language-processing
corpus
information-extraction
english
dataset
named-entity-recognition
corpus-data
ner
relation-extraction
coreference-resolution
-
Updated
Sep 17, 2024 - XML