APPROACH TO AUTOMATED STRUCTURING OF EDUCATIONAL RESOURCES BASED ON THE METHOD OF RECURSIVE REDUCTION

Abstract

The article is devoted to the issues of automated structuring of educational information resources, presented in natural language texts, with the purpose of creating interactive documents based on them. The problems of formation, processing, presentation, use and integration of large volumes of interdisciplinary knowledge are considered. Prospects for the development of information technologies designed to create a formalized system of knowledge in specific subject areas are substantiated. The article analyzes the most common analytical systems that are designed to process natural language texts and can be used to solve this problem. The article proposes a way to evaluate software based on multicriteria optimization. The criteria for such an assessment are identified and justified and its results are presented. Based on the obtained results, it is shown that the most effective tool for solving the problem is the system “Recursive reducer”, which allows to analyze a wide range of documents and present the results in various forms, including XML ontologies that can be used in the work of cognitive IT technology (CIT) “Polyhedron”. The article gives a brief description of the mathematical model on which the “Recursive reducer” based on the use of λ-expressions works. λ-expressions are used to form specialized recursive functions that perform the tasks of identifying tokens in the input text and forming objects. The use of an ontological document classifier to improve the efficiency of a recursive reducer is further described. Article also gives an example of it’s usage — automated structuring of educational information resource, presented by the curriculum in chemistry (advanced training), its presentation in the form of an ontology and creation on its basis of an interactive document by means of CIT “Polyhedron”.