Doroshenko A. Information Technology of Intellectual Analysis of the Fact-based Text Resourses

The actual scientific and practical task of developing models and information tech-nology of intellectual analysis of factual information is solved in the dissertation. On the basis of analysis of models and methods of processing factual data in network streams, the basic requirements for the development of information technology of intellectual analysis of factual resources are formulated. The theory of categories, its projective and predicate interpretations is determined as a mathematical tool for modeling facts. It is proposed to use the theory of intelligence, the method of comparative identification and the apparatus of algebra-logical equations to describe factual information. Models of thematic search and extraction of factual information on the basis of the intellectual procedure for evaluat-ing textual information have been developed. It is proposed to describe the use of two types of triplets: "Subject -Predicate -Object" and "Item -Attribute -Value", which allows you to remove the concept of weakly structured text resources and describe the relationship between them in a structured form. An approach to extracting factual data from text sources has been formed, and the use of ontologies for the description of the processes of integration of factual information is proposed. The use of a new semi-automatic method is proposed for extending the basic ontology, on the example of the subject areas "radiation safety" and "processing of patent information". Approbation of developed models, approaches and information technology was carried out and the results of research were implemented in real information systems. The reference architecture, software components of the server part of the software system, which allows data extraction based on the use of flexible configuration and predicate data mining model, is developed.