Semantic Analysis of Texts Presented in Natural Language

The article considers a cognitive approach to one of the problems of artificial intelligence consisting in semantic analysis of texts presented in natural language. Development of methods and means of discourse analysis is a necessary step for building systems of intellectual communication between a person and a computer (robot). A scientific hypothesis is proposed suggesting that an observer of the environment studies and understands the World when he is able to highlight and analyze Facts, Events and Situations. Corresponding linguistic and cognitive structures of these "units of semantic analysis" of the reality - "Fact", "Event", "Situation" and "Sense" - are proposed and justified. The Sense is considered a prerequisite for understanding dialogue and discourse.Key words: artificial intelligence, cognitive science, knowledge, sense, understanding.Pages of the article in the issue: 228-233Language of the article: Ukrainia