ACL SIGLEX Resource Links

Special Interest Group on the Lexicon of the Association for Computational Linguistics

Corpora

Corpora are intended to be representative of some specified population or genre. Corpora are needed for large scale, systematic contrasts of, for example, language varieties, genres, and modalities (e.g., American vs. British English, informative vs. imaginative prose, or spoken vs. written language). Other research requires enormous amounts of data, even if from fewer genres, as for example, in lexicography, in order to detect words and collocations which occur only rarely. These are recognized corpora used in research; this list does not attempt include all text collections. Also see parallel corpora (involving aligned texts of two or more languages).

Languages

• English
• Dutch
• Portuguese

English

• ICAME Corpora: Corpora available from ICAME (International Computer Archive of Modern and Medieval English) in Bergen, Norway.
  • The Lancaster/Oslo-Bergen Corpus (LOB): Approximately 1,000,000 words of British written English dating from 1960. The corpus is made up of 15 different genre categories. Available as orthographic text and tagged with the CLAWS1 part-of-speech tagging system. The Leeds-Lancaster Treebank and Lancaster Parsed Corpus are analyzed subsamples of the LOB corpus. (See also SIGLEX Treebanks.)
  • The Brown University Corpus: Approximately 1,000,000 words of American written English dating from 1960. The genre categories are parallel to those of the LOB corpus.
  • The Helsinki Corpus (Diachronic Part): samples from texts covering the Old, Middle, and Early Modern English periods. 1,500,000 words in total.
  • Melbourne-Surrey Corpus: 100,000 words of Australian newspaper texts.
  • The Kolhapur Corpus: Approximately 1,000,000 words of Indian written English dating from 1978. Again, the genre categories are parallel to those of the LOB corpus.
  • The Helsinki Corpus of Older Scots: 830,000 words from 1450 - 1700 covering 15 prose genres.
• The British National Corpus, extracts from 4124 modern British English texts of all kinds, both spoken and written, available for academic research to those located in a member state of the EU. Each text is segmented into orthographic sentence units, and each word automatically assigned a part of speech code. There are 6 and a quarter million sentences, and over 100 million words.
• The Cambridge International Corpus at http://uk.cambridge.org/elt/corpus.
• The International Corpus of English (ICE) is collecting corpora in 20 countries and regions (Australia to Zambia). Each participating country is collecting, computerizing, and analyzing a corpus of one million words of their own national or regional variety of English, spoken or written between 1990 and 1996. Each team is following a common corpus design, as well as a common scheme for grammatical annotation. ICE incorporates ICLE, the International Corpus of Learner English.
• For information (academic queries) on the Wellington Corpus of Written New Zealand English (parallel to LOB), the New Zealand ICE Corpus, the Wellington Corpus of Spoken New Zealand English (1 million words) and other smaller collections of data, covering a number of styles, and speakers of different ethnicities, genders, ages, etc., allowing lexical, grammatical and phonetic analysis, please contact Laurie Bauer or Paul Warren.
• The Longman-Lancaster Corpus: Approximately 14.5 million words of written English from various geographical locations in the English-speaking world and of various dates and text types.
• Moby Shakespeare: The complete unabridged works of Shakespeare mshak.tar.Z [2.3MB]

Dutch

• INSTITUUT VOOR NEDERLANDSE LEXICOLOGIE: On-line access to INL 38 Million Words Text Corpus of Dutch, for non-commercial purposes to a Dutch text corpus of ca. 38 million words. Also, from 1994 and 1995, a 5 Million Words Corpus with diversified composition and a 27 Million Words Newspaper Corpus. Access is for free for non-commercial purposes. For access to the corpora, an individual user agreement is required for each corpus. An electronic user agreement form can be obtained from Mailserv@Rulxho.Leidenuniv.NL. Type in the body of your e-mail message:

   	SEND [38MLN96]AGREEMNT.USE  for the 38 Million Words Corpus 1996
   	SEND [27MLN95]AGREEMNT.USE  for the 27 Million Words Newspaper Corpus  1995
   	SEND [5MLN94]AGREEMNT.USE   for the 5 Million Words Corpus 1994
  

  Please make a hard copy of the agreement form, sign it, keep a copy yourself, and return a signed copy to: Institute for Dutch Lexicology INL, P.O. Box 9515, 2300 RA Leiden, The Netherlands. Fax: 31 71 527 2115. After receipt of the signed user agreement, you will be informed about your username and password. If you need additional information, please send an e-mail message to Helpdesk@Rulxho.Leidenuniv.NL, or send a fax to Mrs. dr. J.G. Kruyt.

Portuguese

• The Corpus do Português allows you to quickly and easily search more than 45 million words in more than 50,000 Portuguese texts from the 1300s to the 1900s. The interface allows you to search for exact words or phrases, substrings, lemmas, part of speech, or any combinations of these. You can also search for surrounding words (collocates) within a ten-word window. he corpus also allows you to easily compare (and see, via charts) the frequency of and distribution of words, phrases, and grammatical constructions across texts, in at least three ways:
  • -- By register: comparisons between spoken, fiction, newspaper, and academic
  • -- By dialect: Portugal compared with Brazil
  • -- By historical period: compare different centuries from the 1300s to the 1900s
  You can also easily carry out semantically-based queries of the corpus. For example, you can compare and contrast the collocates of two related words to determine the difference in meaning between these words. You can find the frequency and distribution of synonyms for more than 20,000 words and also compare their frequency in different registers, countries, and historical periods, and use these word lists as part of other queries. Finally, you can easily create your own lists of semantically-related words, and then use them directly as part of the query.

To SIGLEX Resources Main Page