Method for Determining the Level of Criticality Elements when Ensuring the Functional Stability of the System based on Role Analysis of Elements Hryhorii Hnatiienko1, Vladyslav Hnatiienko1, Ravshanbek Zulunov2, Tetiana Babenko1, 3, and Larysa Myrutenko1 1 Taras Shevchenko National University of Kyiv, 64/13 Volodymyrska str., Kyiv, 01601, Ukraine 2 Tashkent University of Information Technologies Ferghana Branch, 108 Amir Temur ave., Tashkent, 100084, Uzbekistan 3 International Information Technology University, 34/1 Manas str., Almaty, A15M0E6, Kazakhstan Abstract The introduction of artificial intelligence technologies in the education process has become an urgent need at the current pace of development of society. The integration of various intelligent technologies is a key factor in this era. The article deals with the issues of adapting the educational process to new technologies. The technology of testing respondents using closed questions focused on multiple-choice answer options is proposed. The paper proposes a new approach to calculating the grade during testing using closed questions oriented to multiple choices. The approaches used earlier in practice were proposed primarily because of their simplicity. However, in connection with the development of soft computing, approaches that were previously used in practice can be supplemented and one should distinguish, for example, a completely incorrect answer from a partially incorrect one. Keywords 1 Artificial intelligence, big data, cloud computing, Internet of Things, intellectual systems, knowledge testing tasks, closed questions. 1. Introduction Society 5.0 is “a human-centered society that balances economic development with the The first industrial revolution is associated with solution of social problems through a highly the development of light industry, the second integrated system of cyberspace and physical (Industrial Society) with the advent of heavy and space.” chemical industries, and the third (Information Society) with the introduction of computers and 2. Directions of Research in the the Internet. The fourth industrial revolution Field of IT implements various technologies such as artificial intelligence, big data, cloud computing, The IoT provides cyber connectivity. Without and the Internet of Things (IoT) [1]. the Internet and an intelligent server system, The integration of various intelligent the IoT is limited to just sensors and actuators technologies is a key factor in this era. The data is [3, 4]. Support for artificial intelligence with really important. Information and data make machine learning, the use of big data allows decisions during this period, and a person or you to process data better and faster, extract it, society must prepare to meet him [2]. In Japan, and make decisions. Intelligent decision this is known as the concept of Society 5.0. support systems are used in a variety of CPITS-2024: Cybersecurity Providing in Information and Telecommunication Systems, February 28, 2024, Kyiv, Ukraine EMAIL: g.gna5@ukr.net (H. Hnatiienko); vladgnat1483@gmail.com (V. Hnatiienko); zulunovrm@gmail.com (R. Zulunov); babenkot@ua.fm (T. Babenko); myrutenko.lara@gmail.com (L. Myrutenko) ORCID: 0000-0002-0465-5018 (H. Hnatiienko); 0009-0000-2678-5158 (V. Hnatiienko); 0000-0002-2132-0834 (R. Zulunov); 0000-0003- 1184-9483 (T. Babenko); 0000-0003-1686-261X (L. Myrutenko) ©️ 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Workshop Proceedings (CEUR-WS.org) CEUR ceur-ws.org Workshop ISSN 1613-0073 Proceedings 301 applications ranging from tourism, finance, • Reasoning is carried out depending and education. Cloud Computing provides a on the logical type. dynamic infrastructure (Cloud Computing) • High-order logical heuristic search, that provides Artificial Intelligence (AI) state-space search, knowledge solutions without large upfront costs [5]. representation. Digital transformation is changing the way • Expert system, automatic theorem organizations operate and deliver services. proving, and design optimization The use of multiple technologies such as using these approaches [10, 11]. artificial intelligence, machine learning, big 2. Numerical approaches: data, IoT, and cloud computing will provide • Facts are represented by numbers improvements. Using these technologies, an [12, 13]. organization can better describe situations, • Numbers are processed using various and be more flexible in turbulence because it algorithms, mainly artificial neural can better predict and apply the recommended networks now also known as deep strategies for the organization [6]. Business learning [14]. process innovation driven by digitally driven • The number of results is expressed in business process reengineering is a key driver the target form [15, 16]. of digital transformation. AI is one of the main • Other methods such as genetic tools for innovation [7]. algorithms, and fuzzy logic [17, 18]. People—Process—Technology (P—P—T) 3. Cognitive approaches (thinking like a are important when introducing a new human): technology. In many cases the People aspect is • This method imitates the process of ignored, most organizations focus only on the human thinking and memorization. Technology aspects. Staff need more time to • Examples: SOAR, CLARION, ACT-R. develop, as well as more financial support. A non-intelligent system can only find the Employees are critical and determine the answer based on the facts (data) available in success of technology adoption and adoption. the database. The answer must be stored in the Strategic workforce planning should be system. It only produces information from included in technology development as well as data. An intelligent system can perform a the implementation of artificial intelligence. thought process to get answers based on Education is a key factor in employee training learned facts. The inference can be based on a [8, 9]. rule or a data pattern. The automation system mainly consists of data input (sensors), automatic processing, and output (actuators). Not all automation 4. Operational Stages of Machine systems use AI, such as the Elevator, a simple Learning line tracker used by a logistics system. Artificial intelligence can help make the The following main operational stages of automation process smarter, more efficient, machine learning are distinguished: and more accurate. AI-assisted automated • Training Phase: In this phase, the AI processing can be performed using a machine outputs a sample training set. The result learning algorithm. AI can learn from of this step is a “model”. A model is a set examples, such as situations in the of values (neural network weights). environment. • Inference phase: In this phase, AI is used to evaluate, predict, control, and infer 3. Basic Approaches in the Field from the learning model based on the of Artificial Intelligence input test. The inference process is carried out using the “model” created during the training phase. Approaches in artificial intelligence: • Evaluation Phase: During the evaluation 1. Symbolic approaches: process, the algorithm is evaluated using • Facts are expressed in symbols. various criteria. • Characters are transformed into other characters by a set of rules. 302 The following types of learning strategies Information technology and artificial are suggested: intelligence enter many aspects of our lives. AI • In unsupervised training, the training set threatens jobs, but it also creates new consists of input patterns only, data opportunities. AI skills are required for all points are not labeled. Algorithms disciplines, not just technicians. It is necessary organize data and group it based on the to prepare the younger generation for the similarity of input patterns. Usually used future, otherwise, there are more risks than when you don’t know what the result opportunities. A national AI strategy is needed, should be, for example, Adaptive and national AI talent should be part of the Resonance Theorem, Self-Organizing national AI strategy. Map, Hopfield Network. Implementation of AI can be carried out in • Supervised training: The training set primary or secondary schools using consists of an input and an expected appropriate tools and teaching materials. output pattern, i.e. a set of labeled However, the most important part of examples. This strategy is used when the educational preparation at the national level is result is known. After training, the the training of teachers. It is far more system can perform classification, important to teach students how to think prediction, and prescribing strategies, computationally than how to use a computer. namely Perception, Forward, The data is important. However, many Backpropagation, Deep Learning, and organizations ignore how they manage data. GAN. An AI solution can only be developed based on • Reinforcement learning: Algorithms that data. Therefore, organizations seeking to learn from results and decide the next implement artificial intelligence must first actions. After each action, the algorithm properly manage their data. This is a positive receives feedback that helps it determine effect of artificial intelligence. Learning paths whether the choice made is right, are important for an AI engineer. AI requires neutral, or wrong. This is a good method basic knowledge, students must follow for use in automated systems that need learning trajectories. New professions require to make many small decisions without “old” knowledge and skills. human intervention. The style should be It is necessary to develop a special switched to Normal. laboratory that encourages cooperation between students. This laboratory should be dedicated to solving AI problems. Students 5. AI is at the Service of Humans from different disciplines sit together at a table to solve AI decisions. Group discussion AI can replace a human (replacement). Many collaboration and following directions are key media such as science fiction films see AI in this learning strategies in this lab. The instructors direction. People fear that AI will replace move from one kiosk (group) to another group. humans. The group works at a table, discusses, and AI helps the person (support). This is the displays the results on a large screen that the current situation. AI-powered systems help group members can see. Groups can work on people in their work environment. Most AI different tasks or assignments and discuss programs are inspired by “natural intelligence” them together in class. and are not yet designed to replace it. The Turing test is a way to test artificial intelligence for human abilities. This test is 6. Types of Collaboration That administered in the form of questions and Can Be Used in the Laboratory answers. Based on the Chinese room paradox, the ability to answer all possible questions that The main types of cooperation that can be used a person can answer does not always indicate in the laboratory are: intelligence, but can also indicate the ability to • Parallel arrangement—students receive remember, think, and produce new the same instructions. It looks like a knowledge—this is the ability of the mind. traditional cool model. 303 • Clear order—in this model, each group The use of artificial intelligence in the works on a separate task. The desktop is education system: a semi-private workspace. • Help the person learn at their own pace. • General arrangement—collaboration • Accurate determination of human needs. between groups is possible, and • Practical solutions to chronic problems. discussion takes place in a large class. • Eliminate red tape in schools. AI can automate key activities in education • Do not waste time in vain. such as assessment. Educational programs can • Improving the quality of education. be adapted to the needs of students. AI can • Ensuring comfort for work. point out areas where courses need • For the right decision thanks to fast data improvement. Students can get additional help analysis. from AI tutors. AI-based software can provide • Planning learning according to the useful feedback for students and educators. AI abilities and pace of the students. is changing the way we find information and • Use or select effective teaching methods interact with it. AI can change the role of through educational analysis. teachers. AI can make learning by trial and • Opportunity to practice in small groups error intimidating. AI-driven data can change with effective planning. how schools find, teach, and support students. • Increasing the efficiency of the AI could change where students learn, who individual learning process. teaches them, and how they acquire key skills. Problems in education and their solutions with the help of AI are presented in Table 1. Table 1 Problems in education and their solutions with the help of AI Number Problem AI Solution of order 1 Standardized curricula are not suitable for individual Personalized education. needs. 2 Limited time available for a tutor. Personal virtual teachers. 3 Big several students in class, many questions cannot be Virtual Classroom assistants. answered. 4 Personalized communication is very difficult for a large Chatbot quickly answers administrative questions. number of students. 5 Selecting the best students from applications. AI can select based on criteria using multiple data. 6 Increasing dropout rates. AI Sentiment Analysis. 7 Difficult to analyze the success of learning experiences. Complements existing learning analytics by providing timely insights into student success, challenges, and needs that can be used to shape the learning experience. 8 Difficult to track the other skills. AI develops reliable and accurate metrics to track student progress, including hard-to-measure traits such as creativity and curiosity. 9 Teachers have to deal the clerical administrative work. The AI acts as an intelligent server to perform clerical tasks. However, the final decision remains with the teacher, as human intelligence is still required. 10 Stop and test approach in assessment. AI can perform qualitative analysis, sentiment analysis, and provide personalized and tailored assessments, and provide role play and collaborative projects within the assessment method. 11 Provide new insights that are difficult or impossible to AI can analyze various data sources to correlate and ascertain from traditional assessments visualize them so that the teacher can better understand the students. AI can be used in education in the following • Grading papers and exams using image cases: recognition, computer vision, and • Academic analytical assessment of predictive methods and learning students and schools using an adaptive analytics using datasets. learning method and a personalized • Real-time virtual personal assistant for learning approach. analytical training. 304 • Intelligent automation of educational elimination at the initial stage of materials and processes. learning. • Creation of automatic learning programs • Customize the learning path for each using augmented intelligence, focused student by collecting learning data. on the specific needs of students. • Identify learning situations and apply • Interaction with students and teachers intelligent adaptive intervention to based on artificial intelligence. students. • Support for students with disabilities and health problems through robotics 7. Testing Tasks and virtual reality. • Identifying students at risk of dropout, A fixed test is the same number of questions for helping them reduce dropout and all students. Most tests are currently used in dropout rates. this model [19, 20]. In an adaptive test, each • Learning a foreign language by speech student is asked a separate question, the recognition and analysis, pronunciation questions are determined by preference and correction, and error correction, recommended by the AI, and the question must reducing the percentage of errors by an be adapted to the abilities of the students. As a average of 83%. result of the test, qualitative and quantitative • Strengthening the decision-making data can be processed. With the help of process with the help of AI. artificial intelligence, you can analyze by • Adaptation and personalization of connecting it with other data sources. It is training programs based on the guaranteed that the estimates will be of higher knowledge, interests, and strengths of quality and more extensive [21, 22]. users. The training catalog should be available on • Create customized textbooks for a the knowledge-sharing platform. The student particular school, course, or even group database also stores the learning path, of students. benefits, class schedule, student qualifications, Functions of AI in education: and expected educational career [23, 24]. 1. In control: Students are enrolled in the system based on • Faster administrative tasks that their wishes, tests, and educational goals. The require study time, such as grading AI-based system offers a curriculum that exams and providing feedback. matches their learning goals. The AI also • Helping teachers with decision checks the available time, training schedule, support and data-driven work. workload, etc. After completing the training • Timely and direct work with the process and passing the exam, the system can student. provide a certificate of completion of the 2. Writing instructions: training as an assessment [25, 26]. This • Predict how a student will exceed approach tailors the learning path to individual expectations in projects and needs and goals [27, 28]. exercises, as well as the dropout rate. • Help teachers create an individual 7.1. Application of Knowledge Testing learning plan for each student. Procedure • Allow learning outside the classroom, and support for collaboration. The knowledge testing procedure is used in • Customize learning styles for each various fields of human activity: programming, student based on their personal technology, medicine, psychiatry, education, etc. information. [29, 30]. In particular, control is an important • Analysis of the proposed program element and one of the most important and course material. components in educational activities [31, 32]. 3. In the process of studying: Moreover, pedagogical control simultaneously • Identification of shortcomings in the performs several functions: educational, student’s learning and their diagnostic, evaluation, stimulating, developing, educational, etc. [32, 33]. 305 Testing is a convenient, but ambiguous way are wrong and they make up a subset 𝐴0 , 𝐴0 ⊂ 𝐴, of assessing knowledge [34, 35]. This and 𝐴1 ∪ 𝐴0 = 𝐴. procedure contains many “pitfalls,” elements In addition, we will assume that all the answers of ambiguity, and lack of justification [36, 37]. offered by the test task 𝑎𝑖 ∈ 𝐴, 𝑖 ∈ 𝐼, are of equal There are many opinions regarding the value. expediency of using tests: on the one hand, For many test tasks, this setting is natural tests are considered a means of positively and logical. For example, to choose among the transforming the educational process in the given options of numbers those that are direction of its technology, reducing labor divisors of the given number. There are many intensity and objectivity; on the other hand, the variants of this kind of task. That is, this tests are seen as a means of degrading the role approach takes place in everyday life and the of the teacher, and the test results are task of its formalization during testing is considered insufficiently reliable [38, 39]. urgent. The peculiarity of such problems lies in the fact that they reflect the well-known truth: 7.2. Types of Test Tasks “How many people have so many opinions”. Therefore, the solution must be justified to the Test tasks are traditionally divided into two extent suggested by the logic of its large groups: construction, the evaluation policy determined • Closed-type test tasks. by the test organizers, common sense, etc. • Open type test tasks. In this paper, we will study a closed-type 7.4. Knowledge Assessment Algorithm task—when each question is accompanied by options for answers, from which several It is proposed to apply an algebraic approach to correct ones should be selected. In turn, closed determining the evaluation results, which is tasks with several options for correct answers successfully used in decision-making theory and provide different options for choosing: the application of expert evaluation technologies. • Task with multiple options—choosing With the algebraic approach, formalization one answer option from the given list. involves the calculation and justification of all • A choice: the subject must answer possible answer options. The maximum number “yes”/“no”. of points for a reliably selected subset of options • Determination of correspondence: the is equal to 𝐵. The number of points for the subject is asked to establish the correctly selected element of the subset of 𝐵 correspondence of the elements of two correct answers 𝑏 = 𝑛 . 1 lists. Problems that a priori depend on a subjective • Establishing the correct sequence—to component cannot be solved without using arrange the elements of the list in a heuristics. A heuristic formula for determining certain sequence, that is, to solve the the score for the choice of answer options in the ranking problem. form of a set generated by the respondent's multiple choice: selection of several answer answers is proposed: 𝑉 ⊂ 𝐴. options from the given test option from the list Moreover, the number of elements 𝜈 = |𝑉| of answers. in the set V can be different: from 0 to n, 0 ≤ |𝜈| ≤ 𝑛. We will denote the number of correct 7.3. Setting the Testing Task answers chosen by the respondent and the number of incorrect answers that he identified Let’s consider the formal description of as correct by 𝜈0 , 𝜈1 + 𝜈0 = 𝜈 ≤ 𝑛. Accordingly, multiple choice in closed questions when the number of answer options that are not testing using models and methods of multiple involved in the respondent’s response to the choice based on the axiom of unbiasedness. question is equal to 𝑛 − 𝜈. Let there be a set of answer options 𝑎𝑖 ∈ 𝐴 Heuristics E1. The value of the penalty for and 𝑖 ∈ 𝐼 = {1, … , 𝑛}, the number of which is each non-matching answer is entered, which is equal to 𝑛, 𝑛 = |𝐴|. Part of the answers 𝑛1 , 𝑛1 < equal to 𝑘: 𝑛, are correct and they make up a subset 𝐴1 , 𝐴1 ⊂ • Е1.1—to some reasonable coefficient 𝑘, 𝐴 , and the other part 𝑛0 , 𝑛0 < 𝑛, of the answers, that reflects the subjective perception of 306 the test organizers about the “price of an 𝑐(𝜈1 = 1, 𝜈0 = 𝑛0 ) and worsens the resulting error”, for example, 𝑘 = 2. estimate by one step, i.e. • E1.2—the value of the expression 𝑘 = 𝑐(𝜈1 = 0, 𝜈0 = 1) = 𝑐(𝜈1 = 1, 𝜈0 = 𝑛0 ) (1) 1 + 𝑝0 , 𝑘 = 1 + 𝑝0 , where 𝑝0 —the − 𝑐(𝜈1 = 1, 𝜈0 = 𝑛0 − 1) probability of an incorrect answer. The following situations of determining the • E1.3—the value of some function resulting assessment are calculated in one of the established by experts 𝑘 = 𝑓Е1 (𝑛0 , 𝑝0 ), ways: which depends on the number of • E5.1—descending function: incorrect answers and their probability. for 𝑖 = 2, … , 𝑛0 Heuristics E2. For an incomplete answer, 𝑐(𝜈1 = 0, 𝜈0 = 𝑖 − 1) 𝑐(𝜈1 = 0, 𝜈0 = 𝑖) = (2) that is, when 𝜈 < 𝑛1 , a partial proportional 𝑘 assignment of points is assumed: • E5.2—the situation 𝑐(𝜈1 = 0, 𝜈0 = 𝑛0 ) is • E2.1—according to the ratio of received equivalent to the situation 𝜈1 + 𝜈0 = 𝑛: correct answers 𝜈1 ≤ 𝜈, to the total 𝑐(𝜈1 = 𝑛1 , 𝜈0 = 𝑛0 ) that is, its number of correct answers 𝑛1 . consequence is a zero rating of the • E2.2—the value of some function respondent. established by the experts 𝑓Е2 (𝑛1 , 𝑝1 ), In this case, the scores for different numbers where 𝑝1 —the probability of receiving of incorrect answers (𝜈0 = 1, … , 𝑛0 ) with zero the correct answer. number of correct answers (𝜈1 = 0) are Of course, a partially correct answer can be determined as follows: guessed by the respondent with a higher 𝑐(𝜈1 = 0, 𝜈0 = 𝑖) = 𝑐(𝜈1 = 1, 𝜈0 = 𝑛0 ) − 𝑖 probability, but the points for it are also ∗ 𝑐(𝜈1 = 1, 𝜈0 = 𝑛0 ) (3) 𝑛0 proportionally lower. where 𝑖 = 1, … , 𝑛0. Heuristics E3. For situations when the respondent did not choose any answer (𝜈 = 0) or all answers were marked as correct, that is 8. Prospects for Further Research 𝜈 = 𝑛, the penalty is a zero evaluation of the on Knowledge Testing Problems result—for lack of selectivity: 𝐵 = 0. An important and ambiguous situation is In the future, modifications of the described when the respondent does not identify a single approach and the development of the idea of correct answer. In this case, respondents may multiple closed-type testing should be be given a different number of incorrect investigated: answers. Depending on the policy of planning • Enter estimates or calculate the test tasks and the position of the decision- numerical value of the accuracy of each maker, the specified situation can be described answer or cluster the variants of the and regulated by additional heuristics. correct answers according to Heuristics E4.1. In the absence of correct significance. answers, the score is always zero, regardless of • Into account indicators of the complexity the number of wrong answers. of questions and answers, for example, Heuristics E4.2. With zero correct answers, depending on the number of fewer incorrect answers are preferred over combinations of correct and incorrect more incorrect answers. answer options or other factors. To formalize this heuristic, we will use the • Consider choosing any number of closed lower limit value of the described situation. To questions from a set of answers—when do this, consider the situation 𝑐(𝜈1 = 1, 𝜈0 = the respondent does not know how 𝑛0 ) when the respondent identified one many answers are correct, thus correct answer and all incorrect ones 𝑛0 . complicating the task. According to the described technology, the • Enter answers of different weights or value of the estimate is determined by the different proximity to the ideal, although following heuristics. this contradicts some principles of Heuristics E5. We will assume that the testing. situation 𝑐(𝜈1 = 0, 𝜈0 = 1) follows the situation 307 Take into account the similarity coefficients Implementation of AI can be carried out in of the answers and the standard—according to primary or secondary schools using the algebraic approach. appropriate tools and teaching materials. However, the most important part of 9. Other Areas of Application of AI educational preparation at the national level is the training of teachers. It is far more important to teach students how to think AI can be applied to businesses and computationally than how to use a computer. organizations to make organizational practices Most programmers or students are not more efficient. Owning artificial intelligence is interested in learning mathematics, logic, or an important tool for the future. This need is statistics. They just want to learn how to necessary not only for workers associated with program. Developing AI solutions requires an the computer but also for other areas. Many understanding of mathematics, statistics, and technologies implement artificial intelligence. logic. The excitement around artificial The decision to implement AI is a critical one intelligence is now pushing programmers to for executives or decision-makers. AI talents study mathematics and statistics. However, are huge opportunities for the development of how AI-enabled theoretical learning materials countries. The role of universities and research are provided needs to change. centers should be enhanced. Students can more freely choose the path of their competence. They may be general to the 10. Conclusions entire curriculum but may focus on specific approaches. Students are encouraged to Artificial intelligence, as a set of new advanced participate in extracurricular activities that are technologies, has emerged relatively recently assessed as credit points. Collaboration and is currently developing rapidly. Some between industry and universities supports, researchers consider this area to be the among other things, the development of technology of the future. Like any other new artificial intelligence education. Competence technology, AI has undoubtedly positive standards in data science and artificial characteristics, but at the same time, it carries intelligence are required. obvious and hidden risks, and perhaps even A learning process that can be easily dangers. replicated should be created to create a large AI is designed to create new content based amount of AI talent. It is necessary to develop on input data or rules. Today, it is used in complete teaching materials that will be free various areas of our lives: e-commerce, energy for teachers. Teachers should be trained in and utilities, telecommunications, automotive training courses before using the material. A and transportation, airport chatbots, etc. The national competency standard should be range of AI applications is constantly growing, established before the development of training and the elements of its presence in people’s materials. lives are steadily increasing. Python and R can be used as programming It is clear that in such conditions, there is a languages for AI. Python starts with minimal need for a systematic analysis of the impact of libraries and can be extended with additional AI on society and the identification of potential libraries. There are several libraries for data problems associated with its development and science and artificial intelligence. Integrated further intellectualization. One of the main Development Environment—Tools such as industries already significantly affected by AI Jupyter are available notebook. Python runs on is education and research. It is important to various operating systems as well as the foresee the peculiarities of AI’s impact on the Anaconda package manager. R is an open- existence and development of this industry, to source statistical program. Designed based on identify its advantages and disadvantages, as the language S. Various ready-to-use packages well as threats from its use. (CRAN) are available. R can be used for The negative impact of AI is largely due to statistical calculations. Integrated its use in generating various kinds of content Development Environments—RStudio and that will contribute to the spread of violations RCmdr are available. of the principles of academic integrity. But this 308 threat should not be exaggerated. It will expediency and security, require certainly lead to the emergence of new trends interdisciplinary research at the intersection of in education that will be aimed at minimizing philosophy, psychology, linguistics, ethics, and such violations. In addition, technologies will other sciences. soon be available to determine whether AI has Artificial intelligence has been developed using been used to generate content with the various disciplines such as philosophy, corresponding consequences. mathematics, economics, neuroscience, The positive effect of the development, use, psychology, computing, control theory, as well as and implementation of AI technologies is much linguistics. Fundamentals of mathematics, greater, and the tasks they can be used to solve statistics, logic, and programming play an can be divided into the following groups: important role in the development of AI 1. Generating express reviews of scientific solutions. Natural language processing such as papers at the initial stages of research in chatbots and sentiment analysis requires an new scientific areas. This can help young understanding of linguistics and psychology. The researchers when writing articles and neural network starts with control theory, and dissertations. now that deep learning has become popular, 2. Advisory assistance to teachers in creating most AI solutions are based on this approach. teaching materials and in generating Therefore, it is necessary to move to a more questions for testing, tasks for interdisciplinary approach to education. independent work, etc. during control A positive feature of the proposed knowledge measures. testing approach is the transparency of the rules 3. Assisting teachers in analyzing answers to set a priori by the test organizers, the absence of open-ended questions when checking uncertainty situations during the evaluation control measures and using AI to procedure, and the monotony of the behavior of automatically evaluate students’ work. the function, which reflects the integral 4. Creating adaptive learning platforms for evaluation of the answers. According to the mass online courses with the ability to proposed technology, the determination of the form individual trajectories and resulting assessment is a well-founded and implement personalized learning, which formalized procedure. In addition, the analyzes student data, including their technology allows for further improvement of academic progress, learning style, and the described approach. other factors to create personalized learning materials and recommendations. References 5. Creating virtual assistants that can support students in their learning process. [1] V. Sokolov, et al., Method for Increasing 6. Creation of anti-generative programs for the Various Sources Data Consistency for identifying text written by artificial IoT Sensors, in: IEEE 9th International intelligence. This problem is caused by the Conference on Problems of Infocom- fact that the use of ChatGPT and other munications, Science and Technology tools for writing text using artificial (PICST) (2023) 522–526. doi: intelligence, especially in dissertations, 10.1109/PICST57299.2022.10238518. theses, research articles, scientific reports, [2] G. Luger, Artificial Intelligence: and other documents whose authors have Structures and Strategies for Complex relevant copyrights, may violate the Problem Solving, Fourth Edition principles of academic integrity. Addison-Wesley (2002). An analysis of technology trends shows that [3] Z. Hu, et al., Development and Operation new results periodically emerge that change our Analysis of Spectrum Monitoring understanding of promising and rational Subsystem 2.4–2.5 GHz Range, Data- directions for the development of society and Centric Business and Applications 48 socio-economic systems. AI technologies are (2020) 675–709. doi: 10.1007/978-3- bringing us closer to creating full-fledged 030-43070-2_29 artificial intelligence systems. The development [4] I. Bogachuk, V. Sokolov, V Buriachok, and coexistence of such systems and human Monitoring Subsystem for Wireless civilization, as well as the problems of 309 Systems based on Miniature Spectrum on Multi-Projection Analysis, in: Analyzers, in: V International Scientific International Scientific Symposium and Practical Conference Problems of “Intelligent Solutions” (IntSol-2023) Vol. Infocommunications. Science and 3538 (2023) 23–32. Technology (2018) 581–585. doi: [15] H. Hnatiienko, N. Tmienova, 10.1109/INFOCOMMST.2018.8632151. A. Kruglov, Methods for Determining [5] N. Nilsson, Artificial Intelligence: A New the Group Ranking of Alternatives for Synthesis, Morgan Kaufmann Publishers Incomplete Expert Rankings, (1998). Mathematical Modeling and Simulation [6] V. Zhebka, et al., Optimization of of Systems (MODS’2020), AISC 1265 Machine Learning Method to Improve (2020) 217–226. doi: 10.1007/978-3- the Management Efficiency of 030-58124-4_21. Heterogeneous Telecommunication [16] H. Hnatiienko, et al., Determining the Network, in: Workshop on Cybersecurity Effectiveness of Scientific Research of Providing in Information and Universities Staff, in: Information Telecommunication Systems, vol. 3288 Technology and Interactions Vol. 2833 (2022) 149–155. (2020) 164–176. [7] S. Russell, P. Norvig, Artificial [17] H. Hnatiienko, V. Snytyuk, O. Suprun, Intelligence: A Modern Approach, Application of Decision-Making Methods Second Edition, Prentice-Hall (2003). for Evaluation of Complex Information [8] V. Buriachok, et al., Implementation of System Functioning Quality, in: Active Cybersecurity Education in Information Technologies and Security Ukrainian Higher School, Information Vol. 2318 (2018) 56–65. Technology for Education, Science, and [18] H. Hnatiienko, V. Snytyuk, A Posteriori Technics, vol. 178 (2023) 533–551. Determination of Expert Competence doi:10.1007/978-3-031-35467-0_32. Under Uncertainty, in: Information [9] V. Buriachok, V. Sokolov, Implementa- Technologies and Security Vol. 2577 tion of Active Learning in the Master’s (2019) 82–99. Program on Cybersecurity, Advances in [19] A. Housen, F. Kuiken, I. Vedder, Computer Science for Engineering and Dimensions of L2 Performance and Education II, vol. 938 (2020) 610-624. Proficiency, Complexity, Accuracy, and doi:10.1007/978-3-030-16621-2_57. Fluency in SLA (2012). [10] J. Giarratano, G. Riley, Expert Systems: [20] P. Black, et al., Working Inside the Black Principles and Programming, Third Box: Assessment for Learning in the Edition Brooks/Cole Publishers (1998). Classroom. Phi Delta Kappan 86(1) [11] P. Jackson, Introduction to Expert (2004) 9–21. doi: 10.1177/0031721704 Systems, Third Edition, Addison-Wesley 08600105. (1998). [21] D. Wiliam, More About Formative [12] A. Voloshin, G. Gnatienko, E. Drobot, A Assessment (2008). URL: Method of Indirect Determination of www.kcl.ac.uk/ssp/education/research Intervals of Weight Coefficients of /iccamsfa.html Parameters for Metricized Relations [22] N. Ziegler, Taking Technology to Task: Between Objects, J. Automat. Inf. Sci. Technology-Media (2016). 35(3) (2003) 25–30. doi: [23] M. García Mayo, Investigating Tasks in 10.1615/JAutomatInfScien. v35.i3.30. Formal Language Learning, Multilingual [13] T. Babenko, et al., Modeling of Critical Matters (2007). Nodes in Complex Poorly Structured [24] R. Brooks, S. Tough, Assessment and Organizational Systems, 26th Testing: Making Space for Teaching and International Conference on Information Learning, Institute for Public Policy Society and University Studies Vol. 2915 Research (2006). (2021) 92–101. [25] L. Bachman, Fundamental Considera- [14] S. Bilan, et al., Technology of Selection tions in Language Testing, Oxford and Recognition of Information Objects University Press (1990). on Images of the Earth’s Surface Based 310 [26] R. Batstone, Planning as discourse Applications to Educational Practice, activity: a sociocognitive view, In Psychol. Learn. Motivation-Adv. Res. Planning and Task Performance in a Theor. 55 (2011) 1–36. doi: Second Language 10 (2005) 277–295. 10.1016/B978-0-12-387691-1.00001-6. [27] H. Hnatiienko, et al., Application of [38] Z. Schrank, An Assessment of Student Expert Decision-Making Technologies Perceptions and Responses to Frequent for Fair Evaluation in Testing Problems, Low-Stakes Testing in Introductory in: Information Technologies and Sociology Classes, Teach. Sociol. 44(2) Security Vol. 2859 (2021) 46–60. (2016) 118–127. doi: 10.1177/0092055 [28] C. Boston, The Concept of Formative X1562474. Assessment, Pract. Assess. Res. Eval. 8(9) [39] R. Thomas, M. McDaniel, Testing and (2002) 1–6. doi: 10.7275/kmcq-dj31. Feedback Effects on Front-End Control [29] D. Duran, J. Builes, A. Gamboa, A Over Later Retrieval, J. Exp. Psychol. Knowledge Management Model for Learn. Mem. Cognit. 39 (2013) 437–450. Improving the Software Test Process, doi: 10.1037/a0028886. European Conference on Knowledge Management 18 (2017) 922–929. [30] V. Reka, Educational Ontology and Knowledge Testing, Electron. J. Knowl. Manag. 5(1) (2007) 123–130. [31] D. Thissen, R. Mislevy, Testing Algorithms, Computerised Adaptive Testing (1990) 103–135. [32] F. Zaromb, H. Roediger, The Testing Effect in Free Recall is Associated with Enhanced Organization Processes, Mem. Cognit. 38 (2010) 995–1008. doi: 10.3758/MC.38.8.995. [33] R. Schmidmaier, et al., Using Electronic Flashcards to Promote Learning in Medical Students: Retesting Versus Restudying, Med. Educ. 45 (2011) 1101– 1110. doi: 10.1111/j.1365-2923.2011. 04043.x. [34] R. Baker, Educational Data Mining and Learning Analytics, The Cambridge Handbook of the Learning Sciences (2019). [35] D. Nurakhmetov, Reinforcement Learning Applied to Adaptive Classification Testing, Theoretical and Practical Advances in Computer-based Educational Measurement, MEMA (2019) 325–336. doi: 10.1007/978-3- 030-18480-3_17. [36] L. Xue-Mei, et al., Research and Implementation of Knowledge Management Methods in Software Testing Process, WRI World Congress on Computer Science and Information Engineering 7 (2009) 739–743. doi: 10.1109/CSIE.2009.360. [37] H. Roediger, A. Putnam, M. Smith, Ten Benefits of Testing and Their 311