Criação de um ambiente para o processamento de córpus de Português Histórico

Abstract

Corpora have been increasingly used within the areas of Linguistics and Natural Language Processing. As a result, new and larger corpora have been compiled and processing systems and standards for encoding and interchange of electronic texts have been developed. However, when it comes to compilation of historical corpora, the methodology is different from the ones used to compile corpora of contemporary language. Another drawback is the fact that most corpus processing systems provide few resources for the treatment of historical corpora, although there are numerous corpora of this type. Similarly, the systems for dictionary creation do not satisfactorily meet the needs of historical dictionaries. The present study is part of a larger project – the Historical Dictionary of Brazilian Portuguese (HDBP) – which aims to compile a dictionary on the basis of a corpus of Brazilian Portuguese texts from the sixteenth through the eighteenth centuries (including some texts from early nineteenth century). Here, we present the challenges for processing the corpus of the HDPB project and establish the criteria for creating the entries of a historical dictionary. This study has developed a computational environment for processing the corpus, building glossaries as well as for creating the entries of the HDPB. This system can be easily adapted to the needs and scope of other historical dictionary projects.