EMCOR: a medical corpus for terminological purposes

Abstract

This article presents the bilingual (French-Spanish) comparable corpus EMCOR and how it has been used to create terminology resources to assist in the translation of medical texts. First of all, attention is paid to the features and interest of the chosen thematic sub-domain, that of the group of rare diseases (RDs) included under the name of inborn errors of metabolism (IEMs). This is followed by an explanation of the criteria used to design EMCOR, its compilation methodology and the processing tasks carried out. Finally, we show the various analytical techniques performed to extract the terms used in naming the pathologies chosen, as well as the concept hierarchy and the bilingual glossary created using the corpus. The study forms part of the research on text translation within the field of biomedicine, and more specifically RDs, on which we are currently working. This is a novel line of research in the medical sub-domain based on the creation of multilingual corpora to produce resources for translators, interpreters or scientific and technical writers that can also be put at the disposal of other users, such as health professionals or even patients themselves.