The parallel Polish-Bulgarian-Russian corpus: problems and solutions

Abstract

The parallel Polish-Bulgarian-Russian corpus we are currently developing as part of CLARIN-PL framework will become an essential tool for translators producing both traditional and digital translations. The electronic tools developed within the project facilitate fast search for and retrieval of multilingual equivalents of lexemes, phrases and sentences. Selected sentences and texts have been semantically annotated for the quantification of nomen, time and aspect. Our definition of equivalent stems from the contemporary contrastive linguistics theory. The guiding principle in the construction of the corpus was to proceed from meaning to form; the principle was first introduced in Koseska-Toszewa (2006). During our work on the Polish-Bulgarian-Russian corpus, we have come across a number of issues, which we regard as characteristic of multilingual corpora: (1) the selection and procurement of texts, (2) the development of computer tools used for the construction of the corpus, (3) multilingual equivalence, and (4) semantic annotation. Multilingual corpora have proved to be exceptionally helpful in language teaching, traditional and digital lexicography, as well as traditional and digital translations. The usefulness of multilingual corpora in each of these areas will be demonstrated through example corpus queries.