Building a Conceptual Structure of a Text in the Field of Electric Power Industry

In this paper, a rule-based Natural Language Processing (NLP) parser is applied to a new type of textual data in the power system engineering domain. The algorithm has been designed to increase the efficiency of circuit analysis and to build a comprehensive conceptual structure of the setup. It provides consistency and accuracy in fact extraction while enabling flexibility with structure alterations in a document. An Extensible Markup Language (XML) metagraph representation has been chosen as a basis for recording and illustrating fact triplets. The model performance was evaluated, and high algorithm capacity to solve text semantic parsing tasks was observed.