Developing a Semantic Role Labeling System in Scientific and Technical Texts on Aviation and Astronautics
Ju. I. Butenko, М.L. GALETKA, E. E. Sineva · Automatic Documentation and Mathematical Linguistics · 2022
Abstract The problem of automatic semantic analysis in natural language processing systems is considered. It is shown that annotating semantic roles is a promising method of analysis, which consists in the recognition of predicate-argument structures in each sentence. A program developed for marking semantic roles in Russian texts is described, and 2000 lexical units are marked on the examples of texts on aviation and astronautics. The unique inventory of semantic roles for aerospace domain is formed. Frame semantics method and ontological approach to frame organization in knowledge base are offered as a model of knowledge representation to organize labeled semantic roles.