Valenda N. Methods and models of functional-semantic processing of natural language texts in the systems of artificial intelligence

The work is devoted to the development of methods and models of the analysis of natural language texts on the basis of functional introduction of semantics. The model of formalization of semantics of natural language designs as superpositions of semantic functions which allows quality improvement of systems of automatic processing of texts at multivalued language units analysis is developed in the thesis. The method of transformation of language designs in superposition of semantic functions is developed. The method of the unification is considered that allows the establishment to semantic equivalence of the language designs expressed by semantic functions. A method of the analysis on the basis of attribute translating grammar, has received the further development due to introduction of semantic functions as symbols of action. Application of the developed methods and models in systems of an artificial intellect for improvement of quality of the analysis of texts in systems of natural language processing is considered.