Khairova N. Linguistic technologies of knowledge identification in semistructured text information

The thesis is dedicated to solving a current applied research problem of methodology introduction to automatize semantic processing of semistructured text information based upon dynamic realization of knowledge identification models. The concept of identification of the contents of elements of coherent texts has been framed and realized in formal models of identification of semantic relations of natural language system using algebra of finite predicates and logical networks. Using the method of formalization of semantic paradigmatic relations of semantic units of semistructured text information logical-linguistic models of description of semantic relations of phrase and superphrasal unities concepts have been developed. Staged linguistic technologies of dynamic retrieval of the contents notions of text repository and semantically oriented identification of tonality of multilanguage texts of a certain knowledge domain, that develop and change, which at the stage of semantic processing implement a model of correlation of the concept of the sign of a language unit and insight understanding of an element of a coherent text have been suggested. The increase of semantic analysis effectiveness of natural language processing systems that provide the implementation of knowledge identification methodology related to natural language texts has been proved experimentally in the thesis.