Lack of appropriate terminological and lexicographical resources is one of the major obstacles that translators face when dealing with specialised translation. Virtual corpora can compensate for this shortage of resources as they allow translators to create and
exploit their own terminological material. The task of manually compiling corpora was arduous and tedious years ago, but thanks to Internet, nowadays it is possible to compile corpora semi-automatically. In addition, it has affected the way in which we conceive of corpora today and the web is even seen a corpus supermarket. Both innovations —the semi-automatic compilation of corpora and the conception of web as a supermarket— are part of our study thanks to Sketch Engine, a corpus manager and text analysis software, and its corpus-building tool WebBootCat. The aim of our study is to use both WebBootCat and Sketch Engine to, first, semi-automatically compile and, second, exploit GeneCorp, our bilingual (French-Spanish), comparable virtual corpus on Genetics. The results of our study show that the combination of both Sketch Engine and our GeneCorp corpus is very convenient as it brings four useful functionalities for translators: 1) the terminological functionality to obtain monolexical and polilexical terminological unit lists; 2) concordance analysis functionality to analyse both collocation behaviour and phraseological pattern of terms; 3) monolingual and bilingual syntactic, grammatical, and collocation behaviour analysis functionality; 4) distributional
thesaurus functionality to clarify conceptual doubts. The combination of virtual bilingual corpora and Sketch Engine helps translators find very quickly no-cost high-quality terminological, phraseological, and thematic resources. Its use is proven to be beneficial for translation lecturers and professional translators and/or interpreters.
