Difference between revisions of "Corpora/Top"

From Nordic Language Processing Laboratory
Jump to: navigation, search
(Activity D: Very Large Corpora)
(Activity D: Very Large Corpora)
 
Line 5: Line 5:
 
  * Reuters Corpus: custom license
 
  * Reuters Corpus: custom license
 
  * NANC: LDC-licensed
 
  * NANC: LDC-licensed
  * GigaWords 5: LDC-licensed; newswire;  
+
  * GigaWords 5: LDC-licensed; newswire; 4.676 billion tokens
  * Wikipedia: Wikipedia Extractor (Wikipedia Corpus Builder on-going)
+
  * Wikipedia: Wikipedia Extractor 2.129 billion tokens (Wikipedia Corpus Builder on-going)
 
  * COW: custom crawl; sentence-shuffled
 
  * COW: custom crawl; sentence-shuffled
 
  * Common Crawl: 140 billion
 
  * Common Crawl: 140 billion

Latest revision as of 05:17, 21 March 2017

Activity D: Very Large Corpora

For English, the UiO team has assessed the following:

* Reuters Corpus: custom license
* NANC: LDC-licensed
* GigaWords 5: LDC-licensed; newswire; 4.676 billion tokens
* Wikipedia: Wikipedia Extractor 2.129 billion tokens (Wikipedia Corpus Builder on-going)
* COW: custom crawl; sentence-shuffled
* Common Crawl: 140 billion
* NOW:

So far, we have standardized on CoreNLP for corpus pre-processing, viz. sentence splitting, tokenization, part of speech tagging, and lemmatization. Each corpus is available in two forms: (a) ‘clean’ running text and (b) pre-processed into the CoNLL-like (tab-separated) output format of CoreNLP.