CC BY
18
НАУЧНО-ТЕХНИЧЕСКИИ ВЕСТНИК ИНФОРМАЦИОННЫХ ТЕХНОЛОГИИ, МЕХАНИКИ И ОПТИКИ сентябрь-октябрь 2016 Том 16 № 5 ISSN 2226-1494 http://ntv.i1mo.ru/
SCIENTIFIC AND TECHNICAL JOURNAL OF INFORMATION TECHNOLOGIES, MECHANICS AND OPTICS September-October 2016 Vol. 16 No 5 ISSN 2226-1494 http://ntv.ifmo.ru/en
RESEARCH OF MULTICRITERIAL DECISION-MAKING MODEL FOR EDUCATIONALa INFORMATIONa SYSTEMS V.V. Serbina, A.M. Syrymbayevaa
a JSC "International Information Technology University (IITU)", Almaty, 050400, Republic of Kazakhstan Corresponding author: syrymbaieva@mail.ru Article info
Received 21.05.16, accepted 12.07.16 doi: 10.17586/2226-1494-2016-16-5-946-951 Article in English
For citation: Serbin V.V., Syrymbayeva A.M. Research of multicriterial decision-making model for educational information systems.
Scientific and Technical Journal of Information Technologies, Mechanics and Optics, 2016, vol. 16, no. 5, pp. 946-951. doi: 10.17586/22261494-2016-16-5-946-951
Abstract
Subject of Research. Decision-making model is offered for informational and educational systems. The study of multi-criteria model is carried out taking into account knowledge, reaction and doubt. Method. The model of material proficiency by the user is based on identification of the personal characteristics when operating with the system. As a result of personal characteristics tracking in the system, an image is formed for each user that can be used for identifying his state: knowledge level, proportion of error, handwriting information, etc. During registration the user is passing an input test. Multi-criteria test results are automatically stored in the user's personal database (agent matrix) and accounted for psychological comfort, formation of the next system content, management of knowledge levels, decision-making when working with the system. The proposed method gives a more clear and "transparent situational picture" for objective decision-making. Main Results. Implementation of multi-criteria decision-making model contributes to the quality of distance education. Also, the method makes it possible to reduce the probability of guessing the correct answer, thus increases the objectivity of knowledge level evaluation in diagnostic systems for management of learning process based on remote technologies. Practical Relevance. Obtained theoretical results of the work are used in training systems on the basis of multi-criteria decision models. Thus, the proposed model leads to an increase in the average score of about 0.3-0.4 points and reduces the training time in 1.5 to 2.0 times. Keywords
model, decision-making model, objective assessment, multi-criteria model, level of doubt УДК 82.05.21
ИССЛЕДОВАНИЕ МУЛЬТИКРИТЕРИАЛЬНОЙ МОДЕЛИ ПРИНЯТИЯ РЕШЕНИЙ ДЛЯ ИНФОРМАЦИО ННО-ОБРАЗОВАТЕЛЬНЫХ СИСТЕМ
В.В. Сербин% А.М. Сырымбаева"
a АО «МУИТ», Алматы, 050040, Республика Казахстан Адрес для переписки: syrymbayeva.assel@gmail.com Информация о статье
Поступила в редакцию 21.05.16, принята к печати 12.07.16 doi: 10.17586/2226-1494-2016-16-5-946-951 Язык статьи - английский
Ссылка для цитирования: Сербин В.В., Сырымбаева А.М. Исследование мультикритериальной модели принятия решений для информационно-образовательных систем // Научно-технический вестник информационных технологий, механики и оптики. 2016. Т. 16. № 5. С. 946-951. doi: 10.17586/2226-1494-2016-16-5-946-951
Аннотация
Предмет исследования. Предложена модель принятия решений для информационно-образовательных систем. Проведено исследование многокритериальной модели, которая учитывает знание, реакцию, а также сомнение. Метод. Модель освоения пользователем материала строится на основе идентификации персональных свойств при работе с системой. В результате отслеживания персональных свойств в системе для каждого пользователя формируется образ, который может использоваться как средство идентификации его состояния: уровень знаний, доля ошибки, информационный почерк и т.д. При регистрации пользователь проходит входное тестирование. Многокритериальные результаты тестирования автоматически сохраняются в персональную базу данных пользователя (матрицу агента) и учитываются для психологического комфорта, формирования следующего контента системы, управления уровнями знаний, принятия решения при работе с системой. Предложенный метод дает более ясную и прозрачную ситуационную картину для объективного принятия решения. Основной результат. Внедрение многокритериальной модели принятия решений способствует повышению качества дистанционного образования.
Также метод дает возможность уменьшить вероятность угадывания правильного ответа, что повышает объективность оценки уровня знаний в системах диагностики для управления процессом обучения по дистанционным технологиям. Практическая ценность. Полученные теоретические результаты использованы в системах обучения на основе многокритериальной модели принятия решений. Предлагаемая модель приводит к увеличению среднего балла примерно на 0,3-0,4 балла и сокращает время на обучение в полтора - два раза. Ключевые слова
модель, модель принятия решений, объективное оценивание, многокритериальная модель, уровень сомнения.
Introduction
The aim of the thesis is the development of information and learning system based on multi-criteria decision model for increase of the learning process efficiency and provision of more qualitative e-education services.
Formulated goal required the following tasks: analysis of the existing research and development in the area of design information and training systems; development of multi-criteria decision model for learning management information system; development of new methods and algorithms to assess the level of knowledge; development of structure and architecture of information and training system focused on human-machine interaction; development of information and software training system; implementation of software in educational institutions and checking the effectiveness of information and training system.
Scientific novelties area is the number of new metrics measuring the level of knowledge, in particular, for the first time we have developed the methods for level of doubt measuring in student's knowledge; we have developed and proposed the original model and multi-criteria evaluation method of diagnosing for the student's level of knowledge with the criterion of doubt level by providing more accurate information measurement; we have developed decision rules for training process control system taking into account not only knowledge, but also the levels of response and user confidence; we have designed data logical model information training system based on multi-criteria decision making model; we have designed architecture, software and information support of information and training system based on multi-criteria decision-making model to ensure the effectiveness of the learning process for loan program.
Currently, the requirements for the quality of university students education (qualification level: knowledge, skills, worldview, mind and senses, abilities, personality and character) requires sophisticated testing methods to detect the level of knowledge, taking into account the social and psychological features of the student in order to manage the learning process effectively. The solution to this problem is possible:
- Firstly, by the integration of all types and methods of testing and validation of knowledge (as well as checking ability, skills and outlook);
- Secondly, by automating the testing process, testing the knowledge and skill level (i.e., quality);
- Thirdly, by achieving maximum objectivity of knowledge evaluation.
Therefore, this paper deals with the latter problem and is focused on maximizing an objective measurement of the level of knowledge. The possibility of solving this problem is to use multi-criterion approach, which measures the number of correct answers given by doubt.
Decision-making model
On the basis of multi-criteria assessment model of knowledge we can identify the main characteristics of the organization and control of automated learning process in information and training system [1]. These include:
- The level of knowledge;
- The level of difficulty;
- The level of reaction;
- The level of confidence [2].
The level of knowledge - the level of current results to the user, based on the coefficient K2 [3]. Difficulty - fixed characteristics prescribed by instructor settings.
The level of response - time assessment of the user's actions in response to any impact. The composition of response level includes coefficients K1, K3 and K5 [4].
The level of confidence - probability characteristic, is inversely proportional to the level of doubt. The structure includes a level of confidence coefficients K4 and K6 [5].
The decision on the basis of multi-criteria model is made in accordance with Figure [6]. Decision-making mode information and training system based on the truth table for the four criteria in accordance with Table 1 [7].
Decision making educational element (action, mode of operation, complexity, time) on the basis of the current state of the educational element is achieved on the basis of the truth table of decision-making in accordance with Table 2 [8].
learn
Figure. Multi-criteria decision making model
State Mode
1 2 3 4
0 0 0 0 Training (opens. Mode)
0 0 0 1 Training (opens. Mode)
0 0 1 0 Training (opens. Mode)
0 0 1 1 Training (closed Mode)
0 1 0 0 Training (closed Mode)
0 1 0 1 Training (closed Mode)
0 1 1 0 Training (opens. Mode)
0 1 1 1 Training mode
1 0 0 0 Training mode
1 0 0 1 Adaptive mode
1 0 1 0 Adaptive mode
1 0 1 1 Setting a level of education
1 1 0 0 Adaptive mode
1 1 0 1 Coaching mode
1 1 1 0 Correction mode
1 1 1 1 Mode self
Table 1. The truth table of decision-making model on four criteria (1 - max, 0 - min, 1 - knowledge, 2 - level of difficulty, 3 - level of reaction, 4 - level of assurance)
The metric scale of measuring the state of the educational element: ignorance (0-49%) is the first measured by Z0. Low level of knowledge (50-74%) is the second measured by ZA. The average level of knowledge (75-89%) is the third measured by ZB. The high level of knowledge (90-100%) is the fourth measured by ZC. Low level of reaction (0-74%) is the fifth measured by RA. The average level of reaction (7589%) is the sixth measured by RB. The high level of reaction (90-100%) is the seventh measured by RC. Low level of confidence (0-74%) is the eighth measured by UA. The next is the average level of confidence (7589%) measured by UB. Then the high level of confidence (90-100%) measured by UC [9].
Complexity has three measures. They are: low complexity measured by A, medium complexity measured by B and high complexity measured by C [10].
State Decision of EE
Knowledge Reaction Confidence Complexity Time Action Mode
Z0 RA UA А max jump theory
Z0 RA UB А max learn theory
Z0 RA UC А max learn theory
Z0 RB UA А - learn theory
Z0 RB UB А - learn theory
Z0 RB UC A - learn theory
Z0 RC UA A min learn theory
Z0 RC UB A min learn theory
Z0 RC UC A min learn theory
ZA RA UA А max repeat exercise
ZA RA UB В max repeat -
ZA RA UC В max jump exam, quiz
ZA RB UA A - repeat exercise
ZA RB UB В - repeat -
ZA RB UC В - jump exam, quiz
ZA RC UA А min repeat exercise
ZA RC UB В min repeat -
ZA RC UC В min jump exam, quiz
ZB RA UA A max repeat exercise
ZB RA UB В max repeat -
ZB RA UC C max jump exam, quiz
ZB RB UA A - repeat exercise
ZB RB UB B - repeat -
ZB RB UC C - jump exam, quiz
ZB RC UA А min repeat exercise
ZB RC UB В min repeat -
ZB RC UC С min jump exam, quiz
ZC RA UA B max repeat exercise
ZC RA UB В max repeat -
ZC RA UC C max jump exam, quiz
ZC RB UA B - repeat exercise
ZC RB UB В - repeat -
ZC RB UC C - jump exam, quiz
ZC RC UA В min repeat exercise
ZC RC UB B min repeat -
ZC RC UC C min jump exam, quiz
Table 2. The current state of the educational element and decision
The learning process organization in information-learning system is based on a measure of doubt for control need rules, which formed the knowledge base [11]. Decision-making model generates rules [12]:
1: if (REs knowledge - ignorance and reaction - low, average or high and
confidence - low, average or high) THEN (complexity - low, the learning
mode - the theory, the effect of educational elements - learn);
2: if (REs knowledge - ignorance, low, average or high and reaction - low and confidence - low, average or high), time (time - max);
3: if (REs knowledge - ignorance, low, average or high and the reaction - high and confidence - low, average or high), time (time - min);
4: if (level of knowledge of UE - low. Medium or high and reaction - low, average or high and confidence - high) THEN (action educational element -jump);
5: if (knowledge UE - low, average or high and reaction - low, average or high and confidence - low or average) THEN (action educational element - repeat);
6: if (knowledge UE - low, average or high and reaction - low, average or high and confidence - low) THEN (training mode - exercise);
10:
11:
12:
if (knowledge UE - low, average or high and reaction - low, average or high and confidence - high) THEN (training mode - exam, quiz);
if (knowledge UE - low or average and reaction - low, average or high and confidence - Low) THEN (difficulty - Low);
if (knowledge UE - low and reaction - low, average or high and confidence -average or high) THEN (difficulty - average);
if (the level of knowledge of UE - average and reaction - low, average or high and confidence - average) THEN (difficulty - average);
if (the level of knowledge of UE - average or high and reaction - low, average or high and confidence - high) THEN (complexity - high);
if (knowledge UE - high and reaction - low, average or high and confidence -low or average) THEN (difficulty - average).
Research of multi-criteria decision-making model for educational information systems
Knowledge Reaction Confidence Mode Time Complexity Ed. element
78 80 70 exercise repeat
50 70 77 average
88 90 50 exercise min average
99 49 66 exercise max repeat
7 49 44 theory low learn
67 66 5 exercise max repeat
77 78 44 exercise repeat
88 90 40 exercise min average
66 98 87 min repeat
99 58 49 exercise max repeat
88 40 50 exercise max repeat
40 44 30 theory low learn
40 40 88 theory low learn
55 55 55 exercise max repeat
5 7 8 theory low learn
88 95 88 min high repeat
4 91 5 theory min low learn
99 6 60 exercise max repeat
70 4 6 exercise max repeat
60 66 4 exercise max repeat
88 85 96 exam,quiz high jump
Table 3. Decision-making model
If a student has an Educational element: learn, he/she should learn theory and spend minimum time and complexity must be of low level [13, 14].
For example, the fifth student has characteristics: knowledge is 7%, reaction is 49% and confidence is 44%. The model gives the following results: educational element is "learn", mode is "theory" and complexity is "low".
In the database one data exists with the level of knowledge equal to 88%, level of reaction equal to 85% and level of confidence equal to 96%, which has educational element "jump", mode is "exam" or "quiz" and complexity is high.
That student should pass Exam or Quiz for finishing the course [15].
Conclusion
This paper deals with creation of decision-making model on the basis of measuring the user's level of doubt to control the learning process. The proposed idea makes it possible to reduce the probability of guessing the correct answer for a more objective assessment of knowledge and adapt the learning process on the basis of the knowledge base.
The results obtained in this study can be used for decision-making in the learning management of information and education distance learning system. The practical value of the work lies in the fact that the use of information in the learning systems based on multi-criteria decision making model obtained in the work leads to increase in the average score on the exam as compared to the control groups by about 0.3-0.4 points and reduce the amount of time required for learning about 1.5-2.0 times.
On the basis of mathematical models and information obtained in the work we have created several computer applications: intelligent information and training system "Programming Languages Borland: Pascal &
Delphi", methodical complex "Mechanics. Molecular physics and thermodynamics", algorithmic learning system "Camel", information learning system for programming "Technology design software based on universal component in Delphi", the interactive test suite of information technology.
Литература
1. Serbin V.V., Syrymbayeva A., Tolebayeva K. Multi-criteria decision-making model for information learning system: a critique of the level of doubt // International Journal of eGovernance and Networks. 2015. N 3. P. 56-65.
2. Сербин В.В., Смирнова Ю.Г. Психометрический показатель сомнения в компьютерных тестах // Proc. Int. Research and Practice Conf. and I stage of the Championship in physical, mathematical and chemical sciences. London, 2013. P. 43-46.
3. Serbin V. Methodology for measuring the level of users doubts: start of a new theory // European Applied Sciences. 2013. N 1. P. 230-233.
4. Serbin V. Multicriteria metod diagnosis of the intellectual and socialpsychological characteristics of personality // Proc. 2nd Int. Scientific Conference European Applied Sciences: Modern Approaches in Scientific Researches. Stuttgart, Germany, 2013. V. 3. Р. 93-97.
5. Ногин В.Д. Принятие решений в многокритериальной среде: количественный подход. М.: Физматлит, 2002. 176 с.
6. Serbin V. Technology, Methodology for the Creation and Development of Information and Learning Systems. Almaty: AIPET, 2010. 198 p.
7. Виноградов Г.П., Кузнецов В.Н. Моделирование поведения агента с учетом субъективных представлений о ситуации выбора // Искусственный интеллект и принятие решений. 2011. № 3. С. 58-72.
8. Теслинов А.Г. Концептуальное проектирование сложных решений. СПб.: Питер, 2009. 288 с.
9. Hoenig C. The Problem Solving Journey: Your Guide To Making Decisions and Getting Results. Perseus Publishing, 2000. 283 p.
10. Galotti K.M. Making Decisions That Matter: How People Face Important Life Choices. London: Lawrence Erlbaum Associates, 2002.
11. Ranyard R., Crozier W.R., Svenson O. Decision Making: Cognitive Models and Explanations. London-NY: Routledge, 1997.
12. Salas E., Klein G.A. Linking Expertise and Naturalistic Decision Making. Lawrance Erlbaum Associates, 2001. 464 p.
13. Stair R.M., Reynolds G. Principles of Information Systems: A Managerial Approach. Delmar Cengage Learning, 2007. 600 p.
14. Applegate L.M., Austin R.D., McFarlan F.W. Corporate Information Strategy and Management: Text and Cases. 7th ed. McGraw Hill, 2007. 657 p.
15. Robinson D. Aspect-Oriented Programming with the e Verification Language. Morgan Kaufmann, 2007. 265 p.
Авторы
Сербин Василий Валерьевич - кандидат технических наук, профессор РАЕ, заведующий кафедрой, АО «МУИТ», Алматы, 050040, Республика Казахстан, syrymbaieva@mail.ru Сырымбаева Асель Маралбайкызы - магистр, лектор, АО «МУИТ», Алматы, 050040, Республика Казахстан, syrymbayeva.assel@gmail.com
References
1. Serbin V.V., Syrymbayeva A., Tolebayeva K. Multi-criteria decision-making model for information learning system: a critique of the level of doubt. International Journal of eGovernance and Networks, 2015, no. 3, pp. 56-65.
2. Serbin V.V., Smirnova Yu.G. Psychometric indicator of doubt in computer test. Proc. Int. Research and Practice Conf. and I stage of the Championship in physical, mathematical and chemical sciences. London, 2013, pp. 43-46.
3. Serbin V. Methodology for measuring the level of users doubts: start of a new theory. European Applied Sciences, 2013, no. 1, pp. 230-233.
4. Serbin V. Multicriteria metod diagnosis of the intellectual and socialpsychological characteristics of personality. Proc. 2nd Int. Scientific Conference European Applied Sciences: Modern Approaches in Scientific Researches. Stuttgart, Germany, 2013, vol. 3, pp. 93-97.
5. Nogin V.D. Decisions in Multicriteria Environment: A Quantitative Approach. Moscow, Fizmatlit Publ., 2002, 176 p.
6. Serbin V. Technology, Methodology for the Creation and Development of Information and Learning Systems. Almaty, AIPET, 2010, 198 p.
7. Vinogradov G.P., Kuznetsov V.N. Modeling agent's behavior based subjective perceptions on the situation of choice. Scientific and Technical Information Processing, 2011, no. 3, pp. 58-72.
8. Teslinov A.G. Conceptual Design of Complex Solutions. St. Petersburg, Piter Publ., 2009, 288 p. (In Russian)
9. Hoenig C. The Problem Solving Journey: Your Guide To Making Decisions and Getting Results. Perseus Publishing, 2000, 283 p.
10. Galotti K.M. Making Decisions That Matter: How People Face Important Life Choices. London, Lawrence Erlbaum Associates, 2002.
11. Ranyard R., Crozier W.R., Svenson O. Decision Making: Cognitive Models and Explanations. London, NY, Routledge, 1997.
12. Salas E., Klein G.A. Linking Expertise and Naturalistic Decision Making. Lawrance Erlbaum Associates, 2001, 464 p.
13. Stair R.M., Reynolds G. Principles of Information Systems: A Managerial Approach. Delmar Cengage Learning, 2007, 600 p.
14. Applegate L.M., Austin R.D., McFarlan F.W. Corporate Information Strategy and Management: Text and Cases. 7th ed. McGraw Hill, 2007, 657 p.
15. Robinson D. Aspect-Oriented Programming with the e Verification Language. Morgan Kaufmann, 2007, 265 p.
Authors
Vassilyi V. Serbin - PhD, Professor, Head of Chair, JSC
"International Information Technology University (IITU)", Almaty,
050040, Republic of Kazakhstan, syrymbaieva@mail.ru
Assel M. Syrymbayeva - magister student, lecturer, JSC
"International Information Technology University (IITU)", Almaty,
050040, Republic of Kazakhstan, syrymbaieva@mail.ru
