Potential of rhetorical relations for differentiation among specialized texts from different domains in Basque and Spanish

This study presents our research on the potential of using rhetorical relations and superficial marks evidencing them to discriminate among specialized texts of different domains but with a high specialization level, in two very different languages as Basque and Spanish. For our analysis, we employ of the Rhetorical Structure Theory (RST). We compiled a parallel corpus of Spanish-Basque specialized texts that contains two subcorpora of medical and terminological texts. We marked these texts with RST rhetorical relations and we detected the discourse markers that evidence them. Finally, we noted that certain types of rhetorical relations and the amount of used discourse markers allow us to differentiate among specialized texts of different domains in both Spanish and Basque.