OWLGOWLGowlgowlgHTMLTXTcorpusdbpediaenglishllodlodlrec-2014nerLicense Not SpecifiedMar 19, 20152015-03-19
Dataset
KAIST silver standard corpus
KAIST silver standard corpus
About
KAIST silver standard corpus
Availability: Freely Avalable
Usage: Named Entity Recognition
Status:Newly created-finished
Description: We propose a novel method to automatically build named entity corpus based on the DBpedia ontology. Our approach is language independent and easy to be applied to other languages where Wikipedia and DBpedia provided. Our corpus which called KAIST silver standard corpus, includes 6,796,274 sentences and 9,522,298 NEs. NE domain is now focus on just PER, ORG, and LOC. But we are now trying to build DBpedia ontology granularity NE corpus as a future work. You can download freely, and can participate refining corpus in our work flow.
Download & API
Description and Data Accesshtml
(as registered in http://www.resourcebook.eu/shareyourlr/index.php#, but currently offline)
CoNLL NE corpus file - Englishtxt
English corpus file as described in "Younggyun Hahm, Kyungtae Lim, Jungyeul Park, YoungSik Kim, Dosam Hwang and Key-Sun Choi. Named Entity Corpus Construction from Wikipedia and DBpedia".
Format is CoNLL: One word per line, words annotated with POS tags and Named Entity Types. Words and annotations tab separated. Sentence boundaries marked by empty lines.
CoNLL NE corpus file - Koreantxt
Korean corpus file as described in "Younggyun Hahm, Kyungtae Lim, Jungyeul Park, YoungSik Kim, Dosam Hwang and Key-Sun Choi. Named Entity Corpus Construction from Wikipedia and DBpedia".
Format is CoNLL: One word per line, words annotated with POS tags and Named Entity Types. Words and annotations tab separated. Sentence boundaries marked by empty lines.
Sources & license
License
- License Not Specified