The Corpus of Portuguese from Academic Journals

ABSTRACT

The present study aims to describe the challenges faced and solutions found in the compilation of the Corpus de Português Escrito em Periódicos - CoPEP, which contains approximately 40 million words, is balanced between the Brazilian Portuguese and European Portuguese varieties in number of words and covers six large areas of knowledge. Firstly, we will present the context of the creation of CoPEP, namely, the make of an on-line dictionary of Portuguese for university students, to which CoPEP served as the primary source for linguistic evidence extraction. Thus, it was the characteristics of this lexicographic project that informed the design criteria for CoPEP and the consequent decision-making process. Next, we will describe the methodology of data acquisition, with a special focus on the challenges that were faced, and the solutions found. We will conclude with the description of the final compilation phase, which involved procedures for obtaining balance.

Keywords:
multivariate corpus; corpus compilation; academic discourse; Portuguese language