Extracting Physical and Technical Structured Information from Natural Language Documents

The paper presents a practical implementation of the developed method for extracting physical and technical structured information from natural language documents. The novelty of the method lies in the cooperation of deep learning technologies based on a tagged corpus of texts containing descriptions of technical and physical effects and methods of semantic-ontological text analysis. •An unlabeled corpus of natural language texts extracted from a unified knowledge base on physical topics was prepared. A tagged corpus of texts containing physical and technical structured information in the form of descriptions of physical effects and their technical implementations was formed for deep learning. Neural network models keyT5, T5 and Bert were trained to extract physical and technical information.