This resource contains a pre-trained BERT language model trained on the Portuguese language. A BERT-Base cased variant was trained on the BrWaC (Brazilian Web as Corpus), a large Portuguese corpus, for 1,000,000 steps, using whole-word mask. The model is available as artifacts for TensorFlow and...

Perfil Sociolinguístico da Fala Bracarense - POS is a manually verified part-of-speech annotation of the EXMARaLDA transcriptions in "Perfil Sociolinguístico da Fala Bracarense", a Portuguese speech corpus with 90 hours of recorded spontaneous speech, aligned with its transcription in EXMARaLDA f...

Geo-Net-PT 02 is a public Geospatial Ontology of Portugal (see Chaves et al., 2007), a computational resource (see Rodrigues et al., 2006 and Rodrigues, 2009) for applications demanding geographic information about Portugal, and contains 701,209 concepts stored in a GKB system, most of them admin...

Arquivo Dialetal CLUP - POS is a speech corpus with approximately 40 000 tokens (Utterances; spontaneous speech, mainly from Northern Portugal). Orthographic transcription, POS.

This is a data set of Portuguese tweets labeled with the emotion conveyed in the tweet. It was gathered using a methodology similar to the one used for building the Affect in Tweets data set used in the SemEval-2018 Task 1. The data set contains 11219 tweets, each labeled with an emotion (anger,...

A corpus of opinion articles annotated with arguments, following a claim-premise model.

Arquivo Dialetal CLUP - ORTH is a speech corpus approximately with 40 000 tokens (Utterances; spontaneous speech, mainly from Northern Portugal). Orthographic and phonetic transcription.

The PropBankPT (Branco et al., 2012) is a set of sentences annotated with their constituency structure and semantic role tags, composed of 3,406 sentences and 44,598 tokens taken from the Wall Street Journal translated. For the creation of this PropBank we adopted a semi-automatic analysis with...

The TreeBankPT (Branco et al., 2011) is a corpus of syntactic constituency trees of the translated news composed of 3,406 sentences and 44,598 tokens taken from the Wall Street Journal. For the creation of this TreeBank we adopted a semi-automatic analysis with a double-blind annotation followed...

The resource is constituted by 20 thousand entries morpho-syntactically and syntactically encoded, accordingly to the parole common encoding standards.