Extracting Terminological Concept Systems

Terminology is the foundation of any specialised communication. As it serves the acquisition of knowledge and the successful communication in a certain domain, terminological inconsistencies represent one major source of misunderstanding in (multilingual) specialised communication.

Automatic term extraction is currently limited to the extraction of a list of term candidates. However, to explicitly display the relation between terms it is necessary to generate terminological concepts and visualise their semantic relations. The Text2TCS application automatically extracts hierarchical and semantic relations from text to create a terminological concept system (TCS). To this end, it relies on findings from ontology learning and machine learning. Such a TCS is a valuable resource in cross-border coordination of communication (beyond language barriers) and extremely important in terms of crises, such as COVID-19. Thereby, it can be ensured that different parties (e.g., health specialists, politicians, journalists) refer to phenomena consistently using the same words. The final outcome of the Text2TCS project is an easy-to-use extraction application that is made freely available on the European Language Grid, a European-wide platform for language technologies and language resources.