OVERVIEW OF THE APPLICATION OF ARTIFICIAL INTELLIGENCE IN EDUCATION

ABSTRACT

Artificial Intelligence (AI), a multidisciplinary field spanning computer science, cognitive science, philosophy, and engineering, is broadly defined by its ability to mimic human cognitive functions like learning, reasoning, and decision-making. This study explores AI’s conceptual roots, historical development, key subfields, and especially its impact on education.

Focusing on educational applications, the research traces the evolution from early computer-assisted instruction to intelligent tutoring systems that enable personalized learning. Contributions by pioneers such as Jaime Carbonell, Edward Shortliffe, and John R. Anderson highlight AI’s role in assessing student performance, identifying learning gaps, and delivering tailored feedback.

The study concludes that AI not only improves educational outcomes but also transforms traditional teaching, promoting more adaptive and individualized learning. It also underscores the importance of ethical and pedagogical considerations in future AI integration.

АННОТАЦИЯ

Искусственный интеллект (ИИ), междисциплинарная область, охватывающая информатику, когнитивную науку, философию и инженерию, в широком смысле определяется его способностью имитировать когнитивные функции человека, такие как обучение, рассуждение и принятие решений. В этом исследовании изучаются концептуальные корни ИИ, историческое развитие, ключевые подобласти и особенно его влияние на образование.

Сосредоточившись на образовательных приложениях, исследование прослеживает эволюцию от раннего обучения с помощью компьютера до интеллектуальных систем обучения, которые обеспечивают персонализированное обучение. Вклады таких пионеров, как Хайме Карбонелл, Эдвард Шортлифф и Джон Р. Андерсон, подчеркивают роль ИИ в оценке успеваемости учащихся, выявлении пробелов в обучении и предоставлении индивидуальной обратной связи.

Исследование приходит к выводу, что ИИ не только улучшает результаты обучения, но и трансформирует традиционное обучение, способствуя более адаптивному и индивидуализированному обучению. Оно также подчеркивает важность этических и педагогических соображений в будущей интеграции ИИ.

Keywords: Artificial Intelligence, Machine Learning, Intelligent Tutoring Systems, Educational Technology, Deep Learning, Natural Language Processing

Ключевые слова: искусственный интеллект, машинное обучение, интеллектуальные системы обучения, образовательные технологии, глубокое обучение, обработка естественного языка

Introduction

Due to encompassing various scientific disciplines and having a broad field of study, there is no universally accepted definition of artificial intelligence (AI) (NASA, n.d.). However, upon examining the definitions that do exist, it is evident that many of them are formed using similar characteristics of AI. Nabiyev and Erümit define artificial intelligence as a general term for computer-based technology that, without relying on a living organism, is created through artificial methods and is capable of exhibiting human-like behaviors and actions, such as thinking, feeling, reasoning, learning, and decision-making.

According to Dick (2019), artificial intelligence is a field that aims to simulate human intelligence capabilities—such as learning, problem-solving, and decision-making—in machines and systems. The Republic of Turkey's Presidential Digital Transformation Office (n.d.) defines AI as the ability of computers or computer-controlled robots to perform tasks associated with intelligent beings. According to Mahto , artificial intelligence is a multidisciplinary field that aims to automate tasks requiring human intelligence.

John McCarthy, considered the father of artificial intelligence, defined AI at the Dartmouth Conference in 1956 as the science of making machines—particularly computers—exhibit intelligent behavior. In 2007, McCarthy updated this definition to describe AI as the science and engineering of creating intelligent computer programs and intelligent machines in a way that is not limited to methods observed in human biology. The key factor behind McCarthy’s evolving definition over time was the advancements in the field of AI.

In this context, artificial intelligence can be defined as systems such as computers, robots, and machines that encompass various disciplines and possess the ability to imitate human-specific abilities. Although there is no universally accepted definition of AI today, based on the literature, it is generally expected that AI systems have the functions of problem-solving, learning, and communication.

Main part

A search engine refers to an integrated system of software and hardware components designed to retrieve specific information based on user queries. Depending on their operational scope, search engines are generally categorized into three types: those intended for personal computer use, systems tailored for corporate or enterprise-level networks, and large-scale engines designed for searching across the Internet.

Cahit Arf made a significant contribution to the discussions on artificial intelligence in Turkey with his 1959 work titled “Can Machines Think and How Can They Think?”. In this work, he examined the possibility of machines possessing the ability to think from a philosophical perspective and discussed the similarities and differences between the functioning of the human mind and that of machines. Arf argued that machines could possess mental abilities such as using language, performing calculations, and making associations. However, he emphasized that endowing machines with human brain features like aesthetic consciousness and free will would be difficult. Evaluating problem-solving abilities through analog and digital machine designs, Arf pointed out that machines have limited capacity to think beyond their programmed boundaries and adapt to new situations.

In 1961, Newell and Simon developed a program called General Problem Solver, which marked an important step toward modeling human reasoning processes. This program is considered one of the first general problem solvers and contributed greatly to symbolic AI research. In 1966, ELIZA, developed by Joseph Weizenbaum, was one of the early works in the field of natural language processing. ELIZA was a program that could converse with humans within a specific dialogue context and demonstrated the potential of AI in human interaction (Weizenbaum, 1966). In 1969, Minsky and Papert published Perceptrons, in which they offered significant criticism of artificial neural networks. This work revealed the limitations of neural networks and led to a temporary decline in neural network research (Minsky & Papert, 1969).

In 1972, the first humanoid intelligent robot, WABOT-I, was produced in Japan. In 1974, MYCIN, developed by Edward Shortliffe, became a notable example of expert systems. MYCIN was an AI system capable of making medical diagnoses and treatment recommendations. This work is considered a pioneer in showcasing the potential of expert systems and the commercial applications of artificial intelligence.

In the history of AI research, periods known as “AI winters” occurred due to reasons such as reduced funding, slower-than-expected progress, unmet high expectations, and insufficiently developed computer systems. The longest of these periods spanned from 1974–1980 and 1987–1993. In the 1980s, the field of AI regained momentum when the UK began funding projects to compete with Japan in AI research. In 1986, Geoffrey Hinton developed the backpropagation algorithm, which became a significant step in training neural networks. This work laid the foundation for deep learning and helped neural networks regain popularity.

In 1997, IBM's Deep Blue program defeated world chess champion Garry Kasparov, marking an important milestone in showcasing AI’s capabilities. In this match, Kasparov competed against a computer program that could process 200 million chess moves per second and was ultimately defeated. This event became one of the first major examples demonstrating that AI could surpass human abilities in strategic games, reinforcing the idea that computers can outperform humans in certain domains.

From the mid-2010s onward, deep learning algorithms revolutionized AI research. Especially through the use of deep neural networks, significant successes were achieved in areas such as image recognition, speech processing, and natural language processing (NLP). In 2012, the deep learning model AlexNet achieved great success in the ImageNet competition, highlighting the power of AI. This model demonstrated that AI could reach human-level performance in image classification tasks.

In 2016, AlphaGo, developed by Google’s^* DeepMind team, became a landmark in demonstrating AI's human-level performance. AlphaGo succeeded in defeating humans in the game of Go, once again revealing the potential of AI in solving complex problems. GPT-3, developed by OpenAI in 2020, emerged as a major breakthrough in the field of artificial intelligence. Using deep learning algorithms, this language model achieved human-level performance in natural language processing and proved effective in many areas such as text generation and translation.

The historical development process of artificial intelligence from the 1950s to the present is illustrated in Figure 1.

Figure 1. Historical development of artificial intelligence

Subfields of artificial intelligence-Artificial intelligence is a broad and comprehensive field of study that encompasses numerous scientific disciplines. Among these, machine learning, deep learning, natural language processing, artificial neural networks, fuzzy logic, and expert systems hold a significant place in the literature. These subfields of artificial intelligence are explained in detail in this section. In addition to these, there are also many other subfields of AI, including robotics, computer vision, speech processing, genetic algorithms, data mining, and cybernetics.

Figure 2. Some sub-branches of artificial intelligence

In 1976, Edward Shortliffe developed a rule-based expert system called MYCIN for diagnosing infectious diseases. In 1979, William Clancey developed a similar system called GUIDON, designed for student use. These systems established an interaction between the student and the expert system by asking various questions about the patient and providing feedback on differential diagnoses, allowing students to experience the role of a doctor during the learning process (Shute & Psotka, 1994).

In 1978, Brown and Burton developed the Buggy software, designed to identify and interpret students’ errors in simple arithmetic operations. The aim was to systematically reproduce student errors and assist educators in recognizing and categorizing these mistakes. However, the Buggy system was not developed to directly teach students but rather to demonstrate the power of automatically generating various possibilities for incorrect answers. After identifying student errors, it provided immediate feedback to correct misunderstandings and guided students toward correct solution methods (Woolf, 1990, 2019).

In the 1990s and 2000s, there was an increase in AI-based educational applications due to advancements in computer technology and the widespread adoption of the internet (Zawacki-Richter et al., 2019). In particular, learning management systems and online learning platforms enhanced the role of AI in education (Ifenthaler & Yau, 2020). With the development of AI and its growing contribution to the field of education, it has increasingly been integrated into teaching and learning processes for various purposes. While AI was once viewed merely as a tool in different stages of the instructional process, it has recently started to appear as a subject to be taught in school curricula. Ministries of education around the world are now conducting serious studies on the benefits of teaching AI to students (Nabiyev & Erümit, 2023). The importance of this study lies in its role as a reference point for the more systematic and effective use of educational technologies.

Conclusion

This research paper examines the scientific foundations, historical development, subfields, and applications of artificial intelligence (AI) in the field of education in detail. Although there is no universally accepted definition of AI due to its diverse scope and dynamic development, a literature review has shown that its main functions encompass human-like abilities such as problem solving, learning, and communication. The historical development process shows that AI has gradually progressed from simple algorithmic systems to complex learning mechanisms, becoming systems capable of making human-level decisions, and demonstrating language understanding.

In particular, in recent decades, the development of machine learning and deep learning technologies has led to the expansion of the application areas of artificial intelligence. The application of these technologies in the field of education has created new opportunities such as personalized learning, adaptive assessment, and intelligent teaching systems. Unlike classical learning approaches, AI-based systems have begun to provide learning environments tailored to the individual needs of students. These technologies analyze students' errors, intervene in their learning process in real time, and provide individual guidance.

As a result, the use of artificial intelligence technologies in education is not only a technical innovation, but also has the potential to change the philosophy and methodology of teaching. In the future, these technologies are expected to be further improved and spread to a wide range of application areas. In this regard, the role of artificial intelligence in education should be investigated not only from a technological but also from a pedagogical perspective, and ethical and social aspects should also be taken into account. This study creates a solid foundation for understanding the current state of artificial intelligence and its future prospects in education.

Reference:

1. Acar, O. (2020). Yapay zekâ fırsat mı yoksa tehdit mi? İstanbul: Kriter Yayınevi.
2. Agarwal, P. K., Gurjar, J., Agarwal, A. K., & Birla, R. (2015). Application of artificial intelligence for development of intelligent transport system in smart cities. Journal of Traffic and Transportation Engineering, 1(1), 20-30.
3. Aggarwal, C. C. (2021). Neural networks and deep learning. The USA: Springer.
4. Akdeniz, M., & Özdinç, F. (2021). Eğitimde yapay zeka konusunda Türkiye adresli çalışmaların incelenmesi. Van Yüzüncü Yıl Üniversitesi Eğitim Fakültesi Dergisi, 18(1), 912-932.
5. Akgun, S., & Greenhow, C. (2022). Artificial intelligence in education: Addressing ethical challenges in K-12 settings. AI and Ethics, 2, 431–440.
6. Akyel, Y., & Tur, E., (2024). Eğitim Bilimlerinde yapay zekânın potansiyeli ve beklentiler, zorluklar ve gelecek yönelimleri. Kırşehir Eğitim Fakültesi Dergisi, 25(1), 645-711.

*По требованию Роскомнадзора информируем, что иностранное лицо, владеющее информационными ресурсами Google является нарушителем законодательства Российской Федерации – прим. ред.)