What’s Out There
← Back to swiping
Language & textFree tier, paid upgradesIndividual

Corpus of Contemporary American English (COCA)

English-Corpora.org (Mark Davies)

A billion words of American English, balanced across genres from TV scripts to academic journals.

A genre-balanced corpus of American English since 1990 covering spoken transcripts, fiction, magazines, newspapers, academic writing, TV and movie subtitles, blogs, and web pages. Linguists, lexicographers, and language teachers use it to see how words are actually used, in what genres, and how that is shifting.

Access

Cost
Free tier, paid upgradesFree web searches with a limited daily quota after registering; full-text and frequency downloads are sold separately.
License
ProprietaryProvider's terms of use; downloadable full text is licensed for purchase
Sign-up
Free account
How you get it
Through the provider's siteWeb interface, Text (purchased), Frequency lists (purchased)

What's in it

One row is
One token (word) with lemma and part-of-speech tag, inside a dated, genre-tagged text
Covers
United States
Years
1990-2019
Updated
Irregular
Size
About 1 billion words
wordlemmapart of speechgenreyearsource title

Quality

4 of 5

The standard reference corpus for American English, with careful genre balance; full data access requires purchase.

  • Downloadable full texts have words removed at intervals for copyright reasons
  • Free interface caps queries per day
  • Most recent texts date to 2019

Connects to

Questions it could answer

  • Which words are far more common in spoken English than in academic writing?
  • How did the use of 'literally' as an intensifier change between 1990 and 2019?

Similar data sets

  • COCO (Common Objects in Context)

    330,000 everyday photos with every person, dog, and fork outlined by hand.

    Images & videoCOCO Consortium (Microsoft and academic partners)Free

  • English Gigaword Fifth Edition

    Ten billion words of newswire from seven agencies, the classic paid corpus behind early NLP.

    Language & textLinguistic Data Consortium (University of Pennsylvania)Paid

  • Common Corpus

    2.27 trillion tokens of text that is either uncopyrighted or openly licensed, with per-document provenance.

    Language & textPleiasFree

  • MultiCoNER Datasets

    Multilingual named-entity recognition data with hard cases like movie titles and short text.

    Language & textAmazonFree

  • xd Crossword Corpus

    Over 6 million crossword clue-answer pairs, from the project that caught a crossword plagiarist.

    Odditiesxd.saul.pw (Saul Pwanson)Free

  • Cornell Movie-Dialogs Corpus

    304,713 lines of dialogue from 617 movie scripts, tagged by character, gender, and billing.

    Film, TV, & booksCornell University (Cristian Danescu-Niculescu-Mizil and Lillian Lee)Free