Search

When forming an ontology, a heuristic approach to the formation and organization of knowledge is often used. The paper proposes a method of metaontologies in the intellectual analysis of text resources, taking into account the fuzziness and blurring of images of natural language units. The use of metaontology makes it possible to organize the search for the necessary data using various inference systems. In the process of information support, metaontology makes it possible to search for information across a large number of distributed heterogeneous data storages and to solve the problem of synonymy and ambiguity of object semantics. The mechanism of dynamic formation of syntagmatic patterns will reduce the impact of the quality of queries to the knowledge base on the quality of answers, as well as increase the speed of finding answers by training the question-answer system. [ABSTRACT FROM AUTHOR]