Abstract
This paper introduces a knowledge base and a method for the automatic classification of given simple sentences of the Kazakh language based on the construction of their parse trees and ontological models. For this, the structure of each type of sentence is represented by the construction of corresponding parse tree. To create a knowledge base taking into account the semantics of the components of simple sentences, ontological models, which will be combined into a single ontological model of simple sentences, are developed. Then, the queries for the knowledge base are formulated and a reasoner, which automatically classifies simple sentences, is launched. The results of the work will be used to create intelligent question-answer, training, knowledge evaluating and other systems.