Abstract

The article examines the possibilities offered by the selection of terminological units drawn from a corpus of German-Russian texts in the field of social work for the compilation of a glossary of frequent collocations. Qualitative and quantitative principles for the definition of terminological units are described, as are the questions of discourse boundaries, of algorithms for generating a general totality and a sampling frame, and of problems in the selection of relevant terminological units. The corpus approach, as a modern tool for modelling vocabulary resources, is employed to compile the glossary at the stage of the selection of terminological units from the texts.