Corpus of American English
From Wikipedia, the free encyclopedia
The freely-available 360+ million word BYU Corpus of American English is the only large corpus of American English currently available, and the only publicly-available corpus of American English to contain a wide array of texts from a number of genres. In addition, since new texts will be added at least two times each year (20 million new words each year), it will serve as a unique linguistic history of American English since 1990.
Contents[hide] |
[edit] Content
The corpus is composed of more than 360 million words in nearly 150,000 texts, including 20 million words each year from 1990-2007. For each year (and therefore overall, as well), the corpus is evenly divided between the five genres of spoken, fiction, popular magazines, newspapers, and academic journals. The texts come from a variety of sources:
- Spoken: (76+ million words) Transcripts of unscripted conversation from nearly 150 different TV and radio programs.
- Fiction: (70 million words) Short stories and plays from literary magazines, children’s magazines, popular magazines, first chapters of first edition books 1990-present, and movie scripts.
- Popular Magazines: (78+ million words) Nearly 100 different magazines, with a good mix (overall, and by year) between specific domains (news, health, home and gardening, women, financial, religion, sports, etc).
- Newspapers: (73+ million words) Ten newspapers from across the US, with a good mix between different sections of the newspapers, such as local news, opinion, sports, financial, etc.
- Academic Journals: (73+ million words) Nearly 100 different peer-reviewed journals. These were selected to cover the entire range of the Library of Congress classification system.
[edit] Queries
- The interface is the same as the BYU-BNC interface for the 100 million word British National Corpus and 100 million word TIME Magazine corpus (see links below)
- Queries by word, phrase, alternates, substring, part of speech, lemma, synonyms (see below), and customized lists (see below)
- The corpus is tagged by CLAWS, the same tagger that was used for the BNC and the TIME corpus
- Chart listings (totals for all matching forms in each genre or year, 1990-present, as well as for sub-genres) and table listings (frequency for each matching form in each genre or year)
- Full collocates searching (up to ten words left and right of node word)
- Comparisons between genres or time periods (e.g. collocates of 'chair' in fiction or academic, nouns with 'break the [N]' in newspapers or academic, adjectives that occur primarily in sports magazines, or verbs that are more common 2004-2007 than previously)
- One-step comparisons of collocates of related words, to study semantic or cultural differences between words (e.g. comparison of collocates of 'small' and 'little', or 'Democrats' and 'Republicans', or 'men' and 'women', or 'rob' vs 'steal')
- Users can include semantic information from a 60,000 entry thesaurus directly as part of the query syntax (e.g. frequency and distribution of synonyms of 'beautiful', synonyms of 'strong' occurring in fiction but not academic, synonyms of 'clean' + noun ('clean the floor', 'washed the dishes')
- Users can also create their own own 'customized' word lists, and then re-use these as part of subsequent queries (e.g. lists related to a particular semantic category (clothes, foods, emotions), or a user-defined part of speech)
[edit] See also
[edit] External links
- Website for the Corpus of American English
- Comparison with the 22 million word American National Corpus
- Other large online corpora from BYU
[edit] References
- Davies, Mark (2008), "Relational databases as a robust architecture for the analysis of word frequency". In AHRC ICT Methods Network: Expert Seminar on Linguistics: Word Frequency and Keyword Extraction, ed. Dawn Archer. Ashgate..
- Davies, Mark (2005), "The advantage of using relational databases for large corpora: speed, advanced queries, and unlimited annotation". International Journal of Corpus Linguistics 10: 301-28.