Internet Access to Corpora: The AC/DC project

Linguateca

Em português


The AC/DC project stands for Acesso a corpora/Disponibilização de corpora ("acess and availability of corpora"), and is one of the activities of Linguateca, previously the Computational Processing of Portuguese project.

The physical address for this service, launched on 23 September 1999, is /ACDC/.

The underlying corpus management system is CWB (a new version of the IMS corpus workbench).

Main goals

One of the goals of Linguateca is to improve significantly the conditions for NLP of Portuguese, namely by

Corpora currently offered and their rough characterization

For each corpus, we present:

(Table below last changed 04 December 2017)

CorpusSize
(units)
Size
(words)
Size
(sentences)
VarietyShort description
AmostRA-NILC128.20398.6334.931BRAmostRA-NILC: pos-tagged sample of NILC Corpus
ANCIB1.707.7181.257.10983.509BRBrazilian discussion list (moderated) on library science
Avante!8.189.8676.762.549212.183PTArticles from Portuguese party-political newspaper Avante!, 1997-2002
Corpus Brasileiro1.175.568.626990.061.95545.457.774BROne thousand million words of Brazilian Portuguese from several genres
CD HAREM290.001225.76612.558PT BRGolden collection of the First and Second HAREMs
CETEMPúblico240.422.936191.272.2468.128.606PTTwo-paragraph excerpts from a major Portuguese daily newspaper, PÚBLICO, 1991-1998
CHAVE124.095.30697.884.7634.682.363PT BRArticles from major daily newspapers PÚBLICO and Folha de São Paulo, 1994-1995
Colonia6.024.3095.028.716216.858PT BRHistorical corpus of Brazilian amd Portuguese texts from XVIth to XXth centuries
CONDIVport7.159.9165.576.547318.776PT BRArticles from sports newspapers and fashion or health magazines from three periods (1950s, 1970s and 2000s), from the CONDIVport project
CONDIVport2142.594117.0944.397PT BRArticles from daily newspapers from the 2010s, from the CONDIVport project
CoNE922.920675.46231.361PT BRSpam or general e-mail messages
C-Oral-Brasil436.444263.64530.647BRC-Oral-Brasil, Brazilian Portuguese informal speech
DiaCLAV7.722.3736.550.993210.373PTArticles from four Portuguese regional newspapers, Diário de Coimbra, Diário de Leiria, Diário de Aveiro, Viseu Diário
Diáspora TL-PT29.36521.8551.038TLDiaspora TL-PT, interviews of East-Timorese in Portugal
ECI-EBR922.439723.99644.381BRCorpus Borba-Ramsey of Brazilian Portuguese
ECI-EE32.13327.130839PTCall for the EU ESPRIT program
ENPCPUB (parte em português)93.16472.3744.371PT BRTranslated fiction from English, subset of the ENPC corpus
Floresta7.252.3066.046.541327.050PT BRThe Floresta treebank
FrasesPB23.18619.145653BRIndividual sentences in Brazilian Portuguese
FrasesPP20.22216.236676PTIndividual sentences in European Portuguese
Mariano Gago349.844278.90412.216PTTexts by and about José Mariano Gago
Moçambula71.39958.2672.321MOMessages from readers to Mozambican newspapers
Museu da Pessoa2.004.2821.422.43493.554PT BRTranscriptions of oral interviews from Museu da Pessoa
Natura/Minho2.208.2111.748.64469.099PTUnedited version of articles for Diário do Minho, a regional newspaper in Portugal
OBras3.637.1302.655.931152.686BRBrazilian literary works in the public domain
P'lo Norte53.70340.9092.384PTBlogs about Norway written by Portuguese blogers
Português Falado - Documentos Autênticos102.22263.3524.735todasTranscribed interviews from ten different locations where Portuguese is spoken
ReLi188.882153.0028.392BRReLi, corpus of book appraisals
NILC/São Carlos42.914.45232.461.8151.988.621BRVarious texts from the NILC Corpus: newspaper, commercial letters and educational texts
todos juntos1.542.626.3671.279.873.86558.767.999todasAll corpora together
Tycho Brahe3.188.7522.669.268109.508PT BRCorpus Tycho Brahe, historical texts
Vercial19.943.49514.393.797895.644PTPortuguese fiction, 16th to XXth century, from Projecto Vercial
Total3.198.472.7672.648.522.943121.880.503PT BRall corpora

We provide more extensive documentation and information, in Portuguese, about the Corpora, the actual processing and encoding of the several kinds of information present in the corpora (tokenization, sentence separation and annotation).

Actual examples on how to invoke AC/DC can be found in exemplos (in Portuguese) and in a page with some old examples in English.

Also, extensive frequency lists are provided in the frequencies service, per corpus, and in total, and computed for both lemmata and forms, also per PoS. In that page, so far only in Portuguese, you are able both to download them or to search them using regular expressions.

Related projects

The following are related projects also under the scope of Linguateca:

[ Access to the AC/DC corpora | Portuguese main page of Linguateca | English page of Linguateca ]


Last updated: 02 August 2016.
We would like to receive your feedback:
Send questions, comments and suggestions