KNOWLEDGE ONTOLOGY MODEL FOR INTELLIGENT TEXT PROCESSING AND ANALYSIS SYSTEMS

Abstract

The article is devoted to solving the scientific problem of creating a top-level description of a knowledge ontology model for intelligent systems for processing and analyzing texts in natural language, built on the basis of an original component architecture that provides the necessary level of detail in the specifications of the analyzed text information. The relevance of this task is due to the need to develop the theoretical foundations for constructing information models of semantic dependencies within texts in natural language. The author gives definitions to the main terms of the subject area under study. A formalized definition of the problem being solved is presented. The problem of the “information explosion,” which was caused by the exponential growth in the volume of digital information, has led to a situation where up to 95% of the information flow contains unstructured data. In such conditions, the task of creating effective intelligent systems for searching and acquiring knowledge, including intelligent systems for processing and analyzing texts in natural language, becomes extremely urgent. The scientific direction for solving this particular problem is Text Mining (TM) – the excavation of knowledge in text information. As an example of the applied task of using acquired knowledge, this study examines the significant problem of information support for the processes of preventing and/or eliminating the consequences of emergency situations. In this task, the initial data are streams of text messages (news information, reports on the technical condition of man-made objects, information about natural phenomena, etc.) arriving at decision- making centers, and the output is formed by predictive assessments and/or specific instructions regarding the assessment situations and actions taken by certain specialists. One of the reasons hindering the development of intelligent text processing and analysis systems for solving problems of searching, acquiring and using knowledge is the insufficiently high level of models and algorithms efficiency that provide a comprehensive solution to the above-described problems of artificial intelligence, taking into account the peculiarities of semantics and context.