FIELD: data processing.
SUBSTANCE: invention relates to computational linguistics. Disclosed is a device for automated extraction of semantic components from compound sentences of natural language texts in machine translation systems, comprising successively connected linguistic modules of sensationalization, fragmentation, graphematization, morphologization, patterning, semantization and extraction, wherein the output of the linguistic patterning module is connected in series to the first input of the indexing module and the filtering module, which forms the linguistic post-patterning module, wherein one output of the filtering module is connected to the second input of the linguistic indexing module, and the second output of the filtering module is connected to the semantization module.
EFFECT: improved quality of automated extraction of semantic components from texts of any level of complexity in interlingual machine translation systems.
13 cl, 1 dwg
