Інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. Частина 2: Дослідження та реалізація
Information technologies, particularly artificial intelligence methods, involve more and more deeply into all spheres of human activity: science, technology, art, and education. Ukraine also has sufficient potential and needs to develop educational support, which is the subject of this paper. The wo...
Збережено в:
| Дата: | 2023 |
|---|---|
| Автори: | , , |
| Формат: | Стаття |
| Мова: | Англійська |
| Опубліковано: |
The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute"
2023
|
| Теми: | |
| Онлайн доступ: | https://journal.iasa.kpi.ua/article/view/285410 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | System research and information technologies |
| Завантажити файл: |  |
Репозитарії
System research and information technologies| _version_ | 1866302916657152000 |
|---|---|
| author | Kulik, Anatoliy Chukhray, Andrey Havrylenko, Olena |
| author_facet | Kulik, Anatoliy Chukhray, Andrey Havrylenko, Olena |
| author_sort | Kulik, Anatoliy |
| baseUrl_str | http://journal.iasa.kpi.ua/oai |
| collection | OJS |
| datestamp_date | 2023-08-07T15:49:29Z |
| description | Information technologies, particularly artificial intelligence methods, involve more and more deeply into all spheres of human activity: science, technology, art, and education. Ukraine also has sufficient potential and needs to develop educational support, which is the subject of this paper. The work aims to demonstrate the components of information technology for the creation of Intelligent Tutoring Systems (ITS), which are involved in studying various engineering disciplines. The work includes methods of system analysis, mathematical and simulation modeling, technical diagnostics, and artificial intelligence. The proposed models and methods are implemented in ITS prototypes for teaching mathematics, programming, and the automatic control theory. The Intelligent Tutoring Systems were implemented in the educational process of KhAI University and other institutions in Ukraine, Great Britain, Austria, and China. Experimental studies have shown increased student learning success rates using ITS compared to traditional methods. Improved and adapted for computer training methods of technical diagnostics, Bayesian networks, and developed models of algorithmic tasks, the learning process and the learner are valuable from a scientific point of view. In a practical sense, the obtained results can be used to create new specialized ITSs and build an expandable common learning platform integrating the basic disciplines of a specific educational field. |
| doi_str_mv | 10.20535/SRIT.2308-8893.2023.2.03 |
| first_indexed | 2025-07-17T10:28:17Z |
| format | Article |
| fulltext |
A.S. Kulik, A.G. Chukhray, O.V. Havrylenko, 2023
Системні дослідження та інформаційні технології, 2023, № 2 35
UDC 004.8: 004.421.2
DOI: 10.20535/SRIT.2308-8893.2023.2.03
INFORMATION TECHNOLOGY FOR CREATING
INTELLIGENT COMPUTER PROGRAMS FOR TRAINING
IN ALGORITHMIC TASKS.
PART 2: RESEARCH AND IMPLEMENTATION
A.S. KULIK, A.G. CHUKHRAY, O.V. HAVRYLENKO
Abstract. Information technologies, particularly artificial intelligence methods, in-
volve more and more deeply into all spheres of human activity: science, technology,
art, and education. Ukraine also has sufficient potential and needs to develop educa-
tional support, which is the subject of this paper. The work aims to demonstrate the
components of information technology for the creation of Intelligent Tutoring Sys-
tems (ITS), which are involved in studying various engineering disciplines. The
work includes methods of system analysis, mathematical and simulation modeling,
technical diagnostics, and artificial intelligence. The proposed models and methods
are implemented in ITS prototypes for teaching mathematics, programming, and the
automatic control theory. The Intelligent Tutoring Systems were implemented in the
educational process of KhAI University and other institutions in Ukraine, Great
Britain, Austria, and China. Experimental studies have shown increased student
learning success rates using ITS compared to traditional methods. Improved and
adapted for computer training methods of technical diagnostics, Bayesian networks,
and developed models of algorithmic tasks, the learning process and the learner are
valuable from a scientific point of view. In a practical sense, the obtained results can
be used to create new specialized ITSs and build an expandable common learning
platform integrating the basic disciplines of a specific educational field.
Keywords: intelligent tutoring systems, experimental studies, results of implemen-
tation.
INTRODUCTION
The implementation of ITS is one of the highest priority directions of educational
tools evolution. This fact is reasoned by numerous advantages of ITS usage over
the classical approach: adoption for particular student, wide possibilities of virtual
modeling of real objects and processes, decreasing of time and work efforts for
completion and verification of tutoring courses, e-learning facilities, etc.
ITSs are characterized by supporting of inner and outer tutoring loops,
minimal feedback (prompting hints and advices), nonlinear learning path, dy-
namic and customizable knowledge base, and self-learning support [1]. ITS usu-
ally include three main structural elements: a domain model, a student model and
a pedagogical model, but researchers also often incorporate an interface model as
the fourth element. Getting into account structure and functionality complexity of
ITS, we can conclude that development process requires huge efforts and deep
knowledge in programming and the subject area. One of the actual problems re-
gards to providing of adaptive hints in the context of certain subject area. A hint
A.S. Kulik, A.G. Chukhray, O.V. Havrylenko
ISSN 1681–6048 System Research & Information Technologies, 2023, № 2 36
must be generated according to the place and type of student’s mistake and as-
sumes possibility of alpha and beta errors. Moreover, it should provide numerous
rules (constraints) for each type of a student’s mistake. Program implementation
of such as laborious process usually requires a high degree of programming skills.
In well-structured domains where solution of the tutoring task can be modeled as
a tree, it is possible to skip manual rule generation (constraints) and automatically
perform pedagogical decisions.
The goal of the presented work was to develop ITSs for different domains on
the basis of system approach [2; 3], models, methods and tools, which were
described in part 1 of the paper [4] and to carry out their experimental studies.
EXPERIMENTAL STUDIES
Experimental studies of the proposed method [4] were performed using m-ary
trees for the metrics of K. Tai, K. Zhang, D. Shasha – the minimum editing dis-
tance between m-ary trees. The performance of the developed NearestHash
method was compared with the speed of the “naive” method and the method of
Bustos, Navarro, Chavez (BNC).
Fig. 1 shows a comparison of the performance of the NearestHash method
with the performance of the BNC method (basic and modified) by input condi-
tions [4]. In the experiment for 10 and 100 trees, the first 10 trees from the origi-
nal list of trees were searched, and in the experiment with 1000 – the first 50. As
can be concluded from the figures, at the first search the developed method sur-
passes the known and modified known for the first experiment from 10 trees 1.71
and 1.67 times; for the second experiment out of 100 trees – 1.66 and 1.69 times
and for the third experiment out of 1000 trees – 3.8 and 2.3 times.
Consider the formation of the set 2SSim [4] based on bigrams, i.e. 2 q .
Let it is given the strings · ]1 (s0) [ 0 · · [3] 0 · [2] 0 · [1] 0 0 lengthsssss
(s0)] [ 0 · lengths , and · ]1 )( [ · · [3] · [2] · [1] snjlengthsnjsnjsnjsnjsnj
]1 )( [ · snjlengthsnj , where ‘·’ is a concatenation symbol.
The third subtask, which is the set 3SSim [4] formation, is solved using a
method developed in collaboration with A.Yu. Zavgorodny [5]. The ITS data
model and SQL-queries constructed within the method allow to choose adaptively
the next task for the student [6].
In the training mode the calculation task process model includes the peda-
gogical action choice, which is based on the information obtained from the stu-
dent’s model and the current step taken by the student in performing the task j .
The various conditions that precede the choice of pedagogical action are:
1. When performing the task j , the student made the wrong step i . In this
case, there are two options:
a) showing a tip for the student on the basis of activated DM;
b) transition to a simpler task, which contains needed KSC.
2. The task is performed correctly by the student. Then there are two possi-
ble scenarios:
a) transition to the next task of the same class;
b) transition to the task of another class.
Information technology for creating intelligent computer programs for training in algorithmic …
Системні дослідження та інформаційні технології, 2023, № 2 37
All these transitions to the next task can be done in two ways:
a) a task is selected by the program;
b) a task is selected by the student.
In the case of student’s incorrect step i while performing the task j hint on
the basis of activated diagnostic models is shown. Then, in the model of the stu-
dent after BN for the step i is inserted as many BN temporary layers, as incorrect
steps have been made. In this case, if we represent the first layer on the left, such
as the graph ) ,( EG , then all duplicate layers will be isomorphic to the origi-
nal graph, and priori values of the probabilities of mastering KSC will be a poste-
riori values of probabilities for the previous layer.
a
b
c
Fig. 1. Comparison of the three methods performance for set, consisting of a — 10, b —
100, c — 1000 randomly generated trees
A.S. Kulik, A.G. Chukhray, O.V. Havrylenko
ISSN 1681–6048 System Research & Information Technologies, 2023, № 2 38
An automatic transition from the current task with the several KSCs should
be made to a simpler task that contains one of the KSCs deal with the appropriate
SQL query. You can also jump to any task that contains fewer KSCs than the current
task using a single SQL query. And the transition to tasks that contain lower number
of KSC, than in the current task, can be done as a result of another SQL-query.
If a student performs the task correctly transition to the next task of the same
class can be done using the method of generation and calculation of objects. As
working with a class of tasks KSC are the same, then priori values of the master-
ing them probabilities will be a posteriori probabilities for the last completed task
of this class. Thus, the values of the probabilities of mastering the relevant KSC
will increase and reach a certain threshold of “mastery”.
If a student solves the task correctly the new task of another class should be
chosen from two variants:
a) a task with at least one KSC, which is new for the student;
b) a task with KSCs, which already occupied by the student, but is combined
in another way.
The above transitions could be performed using SQL queries.
An alternative learning scenario is a scenario using task clustering. Tasks are
grouped into clusters so that each cluster has similar tasks, and clusters are sorted
by complexity – by the average number of KSCs in cluster tasks or by the average
complexity of tasks, where the complexity of the task is defined as the total com-
plexity of the KSC. In this case, the student begins training with the easiest cluster
and does not leave it until all tasks in this cluster are completed. A tuple
) ,, ,( 21 nxxx is used as a representation of the task for clustering, where n is
the total number of KSCs in the problem domain, }1,0{ ix — a component of
the tuple that indicates the presence or absence of the KSC with number i in the
task. Hemming’s metric are usually used as a metric for comparing different
tuples. One of the known methods, for example k-means, is used for clustering of
tasks. In addition, each task can be matched with a hierarchy of KSC, and each
hierarchy – with its string representation. Then NearestHash is used to cluster
tasks with a special data structure – a system of disjoint sets and metrics K. Tai,
K. Zhang, D. Shasha – the minimum editing distance between trees [7; 8].
Since solution of the considered ATs requires to perform calculations using
formulas and algorithms that cover not one specific task, but a whole class of
tasks, it is necessary to separately organize training in the knowledge of these
formulas and algorithms. In such way the student ability to perform any task from
a given class could be improved. Let’s draw the following analogy. Currently,
neural networks are considered to be one of the most effective models of artificial
intelligence. An artificial neural network is learned from examples, but there is no
guarantee that it will work properly with any source data outside the training
sample. Similarly, when teaching a person only by examples, there is no guaran-
tee that he has mastered a general rule, algorithm or formula that covers all possi-
ble cases.
The program compiled by the student might be exactly the same as the refer-
ence program, for example in the case of small tasks. Then, if the texts of these
two programs match, then correctness of the solution could be stated. If the texts
Information technology for creating intelligent computer programs for training in algorithmic …
Системні дослідження та інформаційні технології, 2023, № 2 39
do not match, then program testing is required. If the student’s program does not
match any of the reference programs, then after its testing there are two possible
cases: a) the program is incorrect, b) the program is similar to the correct one. If
the student’s program has passed all the tests and the same or exceeds the existing
reference programs in productivity, the ITS can make this program as a new op-
timal. As criteria of optimality the memory capacity and conciseness should be
used as well. Thus, the automatic self-learning of ITS has been realized.
If the student’s program does not pass at least one test, the next diagnostic
task is solved – finding a place of discrepancy at the lexical level. For this pur-
pose it is used modified Levenstein metric — the minimum editing distance be-
tween arrays of lexems of two programs — and modified Wagner–Fisher method
to determine this distance as well as the trace of the matrix [8], i.e. the minimum
conversion sequence from one program to another.
For the experiment computer Intel Core I5-3210, 2.5 GHz, RAM – 4 GB has
been used. The comparison program was written in Java and run in the Eclipse
SDK, Version: 3.5.1. As a result of comparison of two programs consisting of 28
bytes, it took 386295 ns. For comparison, when calculating unmodified Leven-
stein distances by the Wagner–Fisher method for files with a size of 28 bytes, the
average calculation time is 443109 ns, for files with a size of 2895 bytes —
135536867 ns 0.1 s, for files of size 7251 bytes — 1250641880 ns 1.25 s,
and for files of size 10701 bytes there was an error due to memory overflow.
To solve the problem of diagnosing at the structural level, two abstract syn-
tax trees (AST) are obtained as a result of parsing the reference program and the
program compiled by the student. The first way to compare AST is to use the
width search method. The second way to compare the AST is to calculate the dis-
tance between the trees by the method of K. Zhang, D. Shasha and the editing
trace, i.e. the minimum conversion sequence at the structural level. Thus nodes of
trees are operands and operators.
If the program compiled by the student does not pass at least one test and the
ITS stores more than one reference program for this task, the NearestHash method
is used to find the reference program closest (at the lexical or syntactic level) to
the student’s program.
ALGORITHMS AND SOFTWARE IMPLEMENTATION
The method and models for diagnosing algorithms built by the student in a
graphical environment was developed. Game models are given by two models: a
model like “biathlon” and a model like “sailing regatta”.
As for the first model, each student performs the calculated AT for a certain
period of time finish startt t . Since the task consists of nsteps, in the terms of
biathlon a step i could be considered as a shooting range i . In addition, for each
wrong step a student should get penalty as additional t to the total time. Thus, a
tutor is able:
1) to arrange competitions in the classroom with academic group;
2) to monitor students time at each shooting range;
3) to determine winners at the end of each championship;
4) to determine the absolute record holders after championships series.
The following simplifications were chosen for the second model:
A.S. Kulik, A.G. Chukhray, O.V. Havrylenko
ISSN 1681–6048 System Research & Information Technologies, 2023, № 2 40
1) the yacht moves along a straight line without course change;
2) no currents;
3) the wind speed is stable;
4) the density of environment is constant.
The student’s task is to synthesize the transfer function of the corrective de-
vice, which provides the best quality indicators when turning the sail and, as a
consequence, when moving the yacht. To determine the effectiveness of the ob-
tained student solution, a step-by-step calculation of the values of velocity over
time, as well as the distance traveled. The regatta is visualized: students can com-
pete with each other and with computer models.
The generalized scenario of the “internal cycle” of the ITS for the calculated
AT is shown in Fig. 2. The scheme of the student’s requests analysis method is
shown in Fig. 3.
The methods of determining the similarity of SQL-queries are: 1q — the
classical method of q-grams; 2q – the method of q-grams with the previous re-
placement of keywords in SQL; 3q – the method of q-grams derived from syn-
tactic trees. The total similarity of the two SQL queries 1qry and 2qry is defined
as
3
1
2,12,1
i
qi
qryqry
total
qryqry SIMSIM [9].
Fig. 2. Generalized diagram of the “internal cycle” of the ITS for the calculated AT
X
X
i:=i+1
[i=1
[j=n]
j:=1][i=2[i≥3
j:=j+1]
j:=j+1]
[i:=0]
Fig. 2. Generalized diagram of the “internal cycle” of the ITS for the calculated AT
Information technology for creating intelligent computer programs for training in algorithmic …
Системні дослідження та інформаційні технології, 2023, № 2 41
Fig. 4 shows the sequence of providing tips for a student. An indication of
the error in the query is provided first. If the student is not able to correct the error
on his own, then a more detailed indication of the error is provided, namely its
location based on a comparison of trees. Next, there is a hint about the KSC that
must be used while SQL query construction. Finally, a similar task is demon-
strated with explanation of the necessary steps for its implementation.
The corresponding software was created for the different disciplines of aircraft
control systems department, particularly for automatic control theory (ACT). Fig. 5
Error
presence
Error
place
KSC deal
with error
Similar task
solving demo
Fig. 4. The sequence of providing tips to the student
[High similarity]
[Low similarity]
[Different trees]
[The same tree]
[The same answer]
[Different answers]
SQL similarity checking
Trees comparison
SQL running on a real DB
Correct solution
New reference query register
Incorrect solution
Hints showing
Next task
Fig. 3. Diagram of the student’s SQL-queries analysis method
Fig. 5. ITS for the LGD method and ITS “Sailing Regatta”
A.S. Kulik, A.G. Chukhray, O.V. Havrylenko
ISSN 1681–6048 System Research & Information Technologies, 2023, № 2 42
shows the screen forms of ITS for the method of Lobachevsky-Greffe-Dandelen
(LGD) and ITS “Sailing Regatta”.
The experiments showed that the minimum reduction in the time of manual
completion of tasks by students due to training with ITS support for the LGD
method was 23.3%, and the maximum – 50%. In this case, all students obtained
proper solutions not only in the learning process with the help of ITS, but also
manually during a post-test.
Figs. 6, 7 shows the screen forms of ITS for the construction of frequency
characteristics of the of automatic control object and ITS for the construction of
transient characteristics of the automatic control system.
Fig. 6. ITS for object frequency characteristics construction
Fig. 7. ITS for the transient characteristics construction
Information technology for creating intelligent computer programs for training in algorithmic …
Системні дослідження та інформаційні технології, 2023, № 2 43
As a result of experimental researches it is established that due to ITS for
construction of frequency characteristics of OAU reduction of time of perform-
ance of a task by students made from 25 to 95.4%. As the conducted experiments
showed, thanks to ITS for construction of transitional characteristics of ACS the
minimum reduction of time at manual performance of a task by students made
15%, and the maximum – more than 36.6%.
Fig. 8 shows the screen forms of virtual laboratory work (VLW) “Experi-
mental determination of the parameters of the transfer functions of the objects of
automatic stabilization”.
Fig. 7. ITS for the transient characteristics construction
Fig. 8. Screen forms examples of VLW from automatic control theory
A.S. Kulik, A.G. Chukhray, O.V. Havrylenko
ISSN 1681–6048 System Research & Information Technologies, 2023, № 2 44
The minimum reduction in the time students performed this laboratory work
due to ITS was 33.3%, and the maximum was 43.3%.
Fig. 9 shows screen forms of software for testing knowledge and skills in
mathematics, created by order of Kharkiv Regional Centre of Education Quality
Estimation, as well as Plymouth Inoovation Centre in Mathematics Teaching
(Great Britain).
The minimum time reduction for the development of the site for Centre of
Education Quality Estimation was 553 hours, and the minimum savings – 8188
pounds, which is confirmed by the relevant act.
Fig. 10 shows the screen forms of ITS for algorithms [10] and ITS for SQL
[11]. In the case of ITS for algorithms, each student out of twenty in the aca-
demic group should perform two tasks during the experiments. As a result of the
experiment:
– in the first task: 10 from 20 persons obtained at once a conditional solu-
tion and 10 obtained at once a full solution. From those who obtained a condi-
tional solution, 8 persons managed to obtain a complete solution due to ITS sup-
port later;
Fig. 9. Software for testing knowledge and skills in mathematics
Information technology for creating intelligent computer programs for training in algorithmic …
Системні дослідження та інформаційні технології, 2023, № 2 45
– in the second task: 7 from 18 persons obtained at once a conditional solu-
tion and 6 – a full solution. From those who obtained a conditional solution, 3
persons obtained a full solution with ITS support later.
Regarding the ITS for the SQL language, the experimental study of this ITS
was conducted on 49 KhAI students. At the first stage, students performed 5 tasks
of entrance testing, for which they were given a grades. Then students were solv-
ing 10 tasks in the studying mode and after that they passed a final test, which
consists of the same tasks as the entrance one. As a result, for each student the
score for the final test exceeds the score for the entrance test. The combined his-
togram for entrance and final assessments is shown in Fig. 11.
Fig. 10. ITS screen forms: a — for algorithms; b — for SQL
a
b
Fig. 11. Histogram for entrance and final students testing in ITS for SQL
A.S. Kulik, A.G. Chukhray, O.V. Havrylenko
ISSN 1681–6048 System Research & Information Technologies, 2023, № 2 46
Experimental studies of the software for the ITS automated creation and run
universal environment were also carried out [12]. The screen form of the envi-
ronment and the histogram with results of students training by means of the created
ITS for the discipline "Basics of systems modeling" are shown in Figs. 12, 13.
The experiment participants were 52 students, who were asked to take an en-
trance test, training and final test from the discipline in the ITS environment. As a
result, each student had increased his grades.
CONCLUSIONS
The second part of the paper presents practical results for the development and
implementation of specific ITSs. There were developed different ITS for specific
domains. The experimental studies allow stating following:
1. Every student who studies by means of ITS has occupied necessary
knowledge, skills and competencies while performing tasks, despite possible mis-
takes made during solving, ignorance or incompetence of a student.
Fig. 12. Universal environment for the ITSs automated creation and translation
Fig. 13. Histogram of the students’ learning results with ITS
Information technology for creating intelligent computer programs for training in algorithmic …
Системні дослідження та інформаційні технології, 2023, № 2 47
2. Task solving time has been reduced at least twice on average.
3. Students got skills not only in certain tasks solving, but also in operating
with whole task classes.
4. The tutor routine load has been significantly reduced.
5. Motivation of students to master new competencies has been increased.
The scientific and practical results of the work were shown in one doctoral
and three candidate dissertations. There are over 100 scientific publications, in-
dexed in the Scopus, reported and approved at scientific conferences in Ukraine,
USA, Germany, Austria, Spain, Poland.
The developed ITSs have been implemented in a number of organizations,
educational institutions and enterprises in Ukraine, Austria, Great Britain and China.
REFERENCES
1. A. Chukhray, Methodology for learning algorithms: monograph. National Aero-
space University “KhAI”, 2017, 336 p.
2. M.Z. Zgurovsky and N.D. Pankratova, System analysis: problems, methodology, ap-
plications. Kyiv: Nauk. opinion, 2011.
3. N.D. Pankratova, “System analysis in the dynamics of diagnosing complex technical
systems,” System Research & Information Technology, no. 4, pp. 33–49, 2008.
4. A. Kulik, A. Chukhray, and O. Havrylenko, “Information technology for creating in-
telligent computer programs for training in algorithmic tasks. Part 1: Mathematical
foundations,” System Research & Information Technologies, no. 4, pp. 27–41, 2021.
5. A. Kulik, A. Chukhray, and A. Zavgorodniy, “Similar strings detecting methods,” in
Proceedings of the East-West Fuzzy Colloquium, Zittau, Germany, IPM, 2005,
pp. 183−191.
6. J.P. Martínez Bastida, O. Havrylenko, and A. Chukhray, “Developing a self-
regulation environment in an open learning model with higher fidelity assessment,”
Communications in Computer and Information Science, vol. 826, pp. 112–131, 2018.
7. K. Zhang and D. Shasha, “Simple fast algorithms for the editing distance between
trees and related problems,” Society for Industrial and Applied Mathematics. Jour-
nal on Computing, no. 18 (6), pp. 1245–1262, 1989.
8. A. Chukhray and O.Havrylenko, “Proximate Objects Probabilistic Searching
Method,” Advances in Intelligent Systems and Computing, 1113 AISC, 2020,
pp. 219–227.
9. A. Chukhray and O. Havrylenko, “The method of student’s query analysis while in-
telligent computer tutoring in SQL,” Radioelectronic and Computer Systems, no. 2
(98), pp. 78–86, 2021. doi: 10.32620/re`ks.2021.2.07.
10. Ie. Vagin, O. Havrylenko, J.P. Martínez Bastida, and A. Chukhray, “Computer Intel-
ligent Tutoring System “SQLTOR,”” ICTERI2019 ICT in Education, Research and
Industrial Applications. Integration, Harmonization and Knowledge Transfer,
pp. 541–546. Available: http://ceur-ws.org/Vol-2387/20190525.pdf
11. D. Gaydachuk, O. Havrylenko, J.P.Martínez Bastida, and A. Chukhray, “Structural
diagnosis method for computer programs developed by trainees,” Proceedings of
15th International Conference on ICT in Education, Research, and Industrial Appli-
cations, 2019, Kherson, Ukraine, pp. 485–490.
12. A. Chukhray and O. Havrylenko, “The engineering skills training process modeling
using dynamic Bayesian nets,” Radioelectronic and Computer Systems, no 2 (98),
pp. 87–96, 2021. doi: 10.32620/reks.2021.2.08.
Recieved 27.12.2022
A.S. Kulik, A.G. Chukhray, O.V. Havrylenko
ISSN 1681–6048 System Research & Information Technologies, 2023, № 2 48
INFORMATION ON THE ARTICLE
Anatoliy S. Kulik, ORCID: 0000-0001-8253-8784, National Aerospace University
“Kharkiv Aviation Institute”, Ukraine, e-mail: anatolykulik@gmail.com
Andrey G. Chukhray, ORCID: 0000-0002-8075-3664, National Aerospace University
“Kharkiv Aviation Institute”, Ukraine, e-mail: achukhray@gmail.com
Olena V. Havrylenko, ORCID: 0000-0001-5227-9742, National Aerospace University
“Kharkiv Aviation Institute”, Ukraine, e-mail: lm77191220@gmail.com
ІНФОРМАЦІЙНА ТЕХНОЛОГІЯ СТВОРЕННЯ ІНТЕЛЕКТУАЛЬНИХ
КОМП’ЮТЕРНИХ ПРОГРАМ ДЛЯ НАВЧАННЯ АЛГОРИТМІЧНИМ
ЗАВДАННЯМ. Частина 2: Дослідження та реалізація / А.С. Кулік, А.Г. Чухрай,
О.В. Гавриленко
Анотація. Інформаційні технології і, зокрема, методи штучного інтелекту де-
далі глибше проникають у всі сфери людської діяльності: науку, техніку, мис-
тецтво та освіту. Україна також має достатній потенціал і надзвичайно потре-
бує засобів підтримки навчання, розробці яких і присвячено цю статтю. Метою
роботи є демонстрація компонентів інформаційної технології створення інте-
лектуальних системи навчання, які залучено до вивчення різних дисциплін
технічного профілю. У роботі використано методи системного аналізу, мате-
матичного та імітаційного моделювання, технічного діагностування, штучного
інтелекту. Запропоновані моделі та методи реалізовано в прототипах ITS для
навчання математиці, програмуванню, теорії автоматичного управління. ITS
було впроваджено в навчальному процесі університету ХАІ, інших установах
України, Великобританії, Австрії, Китаю. Експериментальні дослідження по-
казали підвищення показників успішності навчання студентів за допомогою
ITS у порівнянні з традиційними методами. З наукової точки зору мають цін-
ність покращені і адаптовані для комп’ютерного навчання методи технічної
діагностики, Байєсових мереж та розроблені моделі алгоритмічних задач, про-
цесу навчання та особи, що навчається. У практичному сенсі отримані резуль-
тати можна використовувати для створення нових спеціалізованих ITS, а та-
кож для побудови єдиної розширюваної навчаючої платформи, яка об’єднує
базові дисципліни певної галузі.
Ключові слова: інтелектуальні системи навчання, експериментальні дослі-
дження, результати впровадження.
|
| id | journaliasakpiua-article-285410 |
| institution | System research and information technologies |
| keywords_txt_mv | keywords |
| language | English |
| last_indexed | 2025-07-17T10:28:17Z |
| publishDate | 2023 |
| publisher | The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" |
| record_format | ojs |
| resource_txt_mv | journaliasakpiua/44/e8fa0d4db75dc12350942ca7f4d20944.pdf |
| spelling | journaliasakpiua-article-2854102023-08-07T15:49:29Z Information technology for creating intelligent computer programs for training in algorithmic tasks. Part 2: Research and implementation Інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. Частина 2: Дослідження та реалізація Kulik, Anatoliy Chukhray, Andrey Havrylenko, Olena інтелектуальні системи навчання експериментальні дослідження результати впровадження intelligent tutoring systems experimental studies results of implementation Information technologies, particularly artificial intelligence methods, involve more and more deeply into all spheres of human activity: science, technology, art, and education. Ukraine also has sufficient potential and needs to develop educational support, which is the subject of this paper. The work aims to demonstrate the components of information technology for the creation of Intelligent Tutoring Systems (ITS), which are involved in studying various engineering disciplines. The work includes methods of system analysis, mathematical and simulation modeling, technical diagnostics, and artificial intelligence. The proposed models and methods are implemented in ITS prototypes for teaching mathematics, programming, and the automatic control theory. The Intelligent Tutoring Systems were implemented in the educational process of KhAI University and other institutions in Ukraine, Great Britain, Austria, and China. Experimental studies have shown increased student learning success rates using ITS compared to traditional methods. Improved and adapted for computer training methods of technical diagnostics, Bayesian networks, and developed models of algorithmic tasks, the learning process and the learner are valuable from a scientific point of view. In a practical sense, the obtained results can be used to create new specialized ITSs and build an expandable common learning platform integrating the basic disciplines of a specific educational field. Інформаційні технології і, зокрема, методи штучного інтелекту дедалі глибше проникають у всі сфери людської діяльності: науку, техніку, мистецтво та освіту. Україна також має достатній потенціал і надзвичайно потребує засобів підтримки навчання, розробці яких і присвячено цю статтю. Метою роботи є демонстрація компонентів інформаційної технології створення інтелектуальних системи навчання, які залучено до вивчення різних дисциплін технічного профілю. У роботі використано методи системного аналізу, математичного та імітаційного моделювання, технічного діагностування, штучного інтелекту. Запропоновані моделі та методи реалізовано в прототипах ITS для навчання математиці, програмуванню, теорії автоматичного управління. ITS було впроваджено в навчальному процесі університету ХАІ, інших установах України, Великобританії, Австрії, Китаю. Експериментальні дослідження показали підвищення показників успішності навчання студентів за допомогою ITS у порівнянні з традиційними методами. З наукової точки зору мають цінність покращені і адаптовані для комп’ютерного навчання методи технічної діагностики, Байєсових мереж та розроблені моделі алгоритмічних задач, процесу навчання та особи, що навчається. У практичному сенсі отримані результати можна використовувати для створення нових спеціалізованих ITS, а також для побудови єдиної розширюваної навчаючої платформи, яка об’єднує базові дисципліни певної галузі. The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" 2023-06-30 Article Article application/pdf https://journal.iasa.kpi.ua/article/view/285410 10.20535/SRIT.2308-8893.2023.2.03 System research and information technologies; No. 2 (2023); 35-48 Системные исследования и информационные технологии; № 2 (2023); 35-48 Системні дослідження та інформаційні технології; № 2 (2023); 35-48 2308-8893 1681-6048 en https://journal.iasa.kpi.ua/article/view/285410/279507 |
| spellingShingle | інтелектуальні системи навчання експериментальні дослідження результати впровадження Kulik, Anatoliy Chukhray, Andrey Havrylenko, Olena Інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. Частина 2: Дослідження та реалізація |
| title | Інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. Частина 2: Дослідження та реалізація |
| title_alt | Information technology for creating intelligent computer programs for training in algorithmic tasks. Part 2: Research and implementation |
| title_full | Інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. Частина 2: Дослідження та реалізація |
| title_fullStr | Інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. Частина 2: Дослідження та реалізація |
| title_full_unstemmed | Інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. Частина 2: Дослідження та реалізація |
| title_short | Інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. Частина 2: Дослідження та реалізація |
| title_sort | інформаційна технологія створення інтелектуальних комп’ютерних програм для навчання алгоритмічним завданням. частина 2: дослідження та реалізація |
| topic | інтелектуальні системи навчання експериментальні дослідження результати впровадження |
| topic_facet | інтелектуальні системи навчання експериментальні дослідження результати впровадження intelligent tutoring systems experimental studies results of implementation |
| url | https://journal.iasa.kpi.ua/article/view/285410 |
| work_keys_str_mv | AT kulikanatoliy informationtechnologyforcreatingintelligentcomputerprogramsfortraininginalgorithmictaskspart2researchandimplementation AT chukhrayandrey informationtechnologyforcreatingintelligentcomputerprogramsfortraininginalgorithmictaskspart2researchandimplementation AT havrylenkoolena informationtechnologyforcreatingintelligentcomputerprogramsfortraininginalgorithmictaskspart2researchandimplementation AT kulikanatoliy ínformacíjnatehnologíâstvorennâíntelektualʹnihkompûternihprogramdlânavčannâalgoritmíčnimzavdannâmčastina2doslídžennâtarealízacíâ AT chukhrayandrey ínformacíjnatehnologíâstvorennâíntelektualʹnihkompûternihprogramdlânavčannâalgoritmíčnimzavdannâmčastina2doslídžennâtarealízacíâ AT havrylenkoolena ínformacíjnatehnologíâstvorennâíntelektualʹnihkompûternihprogramdlânavčannâalgoritmíčnimzavdannâmčastina2doslídžennâtarealízacíâ |