KNOWLEDGE BASE ARCHITECTURE OF AUTOMATED EDUCATIONAL SYSTEM

Author:

Posadska Iryna, Chernihiv National University of Technology (95 Shevchenka Str., 14027 Chernihiv, Ukraine)

Language: russian

Annotation:

Urgency of the research. Creation of knowledge base architecture, which will have the ability to work with different types of knowledge representations and keep them up-to-date, is the most important task in designing a knowledge-oriented automated learning system.

Target setting. Knowledge-oriented automated learning systems are one of the types of automated learning systems. After analyzing the features of building the knowledge base architecture of such systems, it became necessary to offer a new architecture that makes it possible to store knowledge in an actual form without affecting the interaction of functional modules of automated learning systems based on knowledge.

Actual scientific researches and issues analysis. A wide range of knowledge of the course domain, in particular natural language presentations of the training course, formalized representations of domain fragments in the form of UML diagrams, formalized representations based on a certain calculus such as the predicate calculus of the first order, frames, product rules and others, are presented in recent studies. However, each species has advantages and disadvantages.

Uninvestigatigated parts of general matters defining. Considered types of knowledge representations about the course domain have significant drawbacks. Therefore, it becomes necessary to propose other criteria for presenting knowledge about the training course, which will be the basis for building the knowledge base architecture of a knowledge-oriented automated learning system.

The research objective. The purpose of the article is to describe and substantiate the knowledge base architecture for a knowledge-oriented automated learning system.

The statement of basic materials. Different types of knowledge representations are analyzed, their advantages and disadvantages are considered. Since there is no best representation of knowledge about the training course, it is suggested that several of its representations should be used in accordance with the purposes of use. Three-layer architecture and features of the knowledge base of knowledge-oriented automated learning systems are proposed. The criterion of the ability to control the assimilation of educational material is highlighted among other criteria for the methods of representation of knowledge. Conducting the surveys on the passed material (examination) is the basic for the control at training. Similarities of the knowledge control process and the testing process of object-oriented programs are considered in the article. Levels of complexity tests also are highlighted.

Conclusions. The proposed architecture allows to save and effectively use the advantages of each of the layers of the knowledge representation in the common system. An important role in ensuring the efficiency of the system is played by the converters «NL text →UML» and «UML → Logical representations» with this architecture.

Key words:

automated learning system, knowledge, knowledge base, formalization, domain

References:

1. Lytvynov, V., Posadska, I. (2015). Knowledge representation in the automated learning systems. International Journal “Information Technologies & Knowledge”, vol. 9, no. 1, pp. 34–43.

2. Litvinov, V.V., Posadskaya, I.S., Savelev, M.V. (2016). Arkhitektura znanie-orientirovannoі avtomatizirovannoі sistemy obucheniіa [Architecture of knowledge-oriented automated system of learning]. Tekhnіchnі nauki ta tekhnologії – Technical Sciences and Technologies, no. 3 (5), pp. 122–130 (in Russian).

3. Klykov, Іu.I., Gorkov, L.N. (1980). Banki dannykh dlia priniatiia reshenii [Databases for decision-making]. Moscow: Sov. radio (in Russian).

4. Shustov, S.B. (2009). Teoriia resursov i resursnye krizisy: proshloe, nastoiashhee i budushhee: analiticheskii obzor [The theory of resources and resource crises: past, present and future: an analytical review]. Nizhnii Novgorod: Nizhnii Novgorod (in Russian).

5. Douglass, Bruce Powel (1999). Real - Time UML. Second Edition. Developing Efficient Objects for Embedded Systems. Wesley.

6. Lorer Zh.-L. (1991). Sistemy iskusstvennogo intellekta [Artificial Intelligence Systems: Translation from French] [Trans. from French]. Moscow: Mir (in Russian).

7. Minsky, Marvin (1975). A Framework for Representing Knowledge: Patrick Henry Winston. The Psychology of Computer Vision. McGraw-Hill: New York (U.S.A.).

8. Popova, E.V. (ed.) (1990). Iskusstvennyi intellekt. Kn. 2: Modeli i metody: spravochnik [Artificial Intelligence. Book 2: Models and Methods: A Handbook]. MoscowRadio i sviaz (in Russian).

9. Piterson, Dzh. (1984). Teoriia setei Petri i modelirovanie sistem [The theory of Petri nets and modeling systems]. Moscow: Mir (in Russian).

10. Korolyuk, V.S. & Turbin, A.F. (1982). Protsessy markovskogo vosstanovleniya v zadachakh nadezhnosti system [Markov renewal processes in systems reliability problems]. Kiev : Nauk. dumka (in Russian).

Download