CC BY
0Эволюция методик адаптивного обучения: аспекты внедрения технологий искусственного интеллекта
Кудинова Татьяна Викторовна,
старший преподаватель, кафедра иностранных языков ИРИ, МИРЭА Российский технологический университет E-mail: kudinova@mirea.ru
В статье рассматривается эволюция адаптивного обучения, его истоки восходят к Древней Греции и персонализированному интерактивному методу Сократа. Этот исторический подход заложил основу для современных практик адаптивного обучения. На протяжении столетий значительные вехи в теории образования, установленные Джоном Амосом Коменским, Львом Выготским и Беррхусом Фредериком Скиннером, постепенно формировали адаптивное обучение. Автор подчеркивает, что технологические достижения конца XX века, особенно в области компьютерного обучения и когнитивного репетиторства, способствовали дальнейшему развитию методов адаптивного обучения. Интеграция искусственного интеллекта и машинного обучения в современные платформы адаптивного обучения произвела революцию в образовании, позволив проводить персонализированное обучение в режиме реального времени. Несмотря на такие проблемы, как сбор данных, сложность и необходимость значительных инвестиций, будущее адаптивного обучения остается многообещающим. Оно способно повысить мотивацию и вовлеченность учащихся, улучшить результаты обучения, повысить трудоспособность и снизить уровень отсева. Тем не менее, проблемы остаются, включая трудности адаптации к широкому спектру дисциплин, измерение сложных характеристик, а также временные и финансовые затраты, необходимые для успешного внедрения.
Ключевые слова: адаптивное обучение, персонализированное обучение, программированное обучение, образовательные технологии, ИИ, машинное обучение.
о с
U
см о см
1Л
Introduction
Adaptive teaching is an educational approach that customizes instruction to meet the individual needs of students. While traditional teaching methods use a universal model, adaptive learning takes into account each student's individual abilities, learning styles, background knowledge and pace of learning. The main objective is to achieve maximum efficiency and productivity of each student's learning by adjusting the content, methods and intensity of learning. Educational technologies that enable this method to be implemented are called adaptive learning technologies.
In the last few years, modern technologies, including artificial intelligence technologies, cloud computing, big data, computer software and mobile devices, play an important role in the implementation of adaptive learning, providing powerful opportunities to tailor and enhance the learning experience for students. Due to these trends, the research and practice of adaptive learning is gaining momentum, hence, it continues to be a growing and popular topic.
American scholar Green considers the use of adaptive learning technologies to be very promising for interdisciplinary research. Gordon Pask, a British cyber-netician and educational theorist, introduced the concept of adaptive learning in the context of his research on cybernetics and educational technology. Pask was sure that educational systems should be self-adaptive, meaning they should be able to modify their instructional strategies based on the learner's responses and progress. This involves continuously assessing the learner's knowledge and skills, and then adjusting the content, difficulty, and mode of instruction accordingly. [8].
Programmed and adaptive learning technologies were formed in 1950-1960 with the participation of such scientists as Gordon Pask, Burrhus Frederic and Norman Crowder. They proposed their own algorithms and technologies to enhance learning through the use of systematic and adaptive approaches. [8, 10, 3]. It is believed that the use of adaptive learning spread in the 70s due to the fact that it was at this time that the latest computer programme SCHOLAR appeared. [2].
There are two fundamentally different schools of adaptive learning: Russian and Western. The theoretical foundations and methodology of these approaches differ significantly. Thus, the Russian approach is based on the theory of gradual mental actions formation, which was elaborated by Russian psychologists Lev Vygotsky and Pyotr Galperin. The Western approach to adaptive learning is based on behaviourism, a theory that emphasises controlled behaviour and ex-
ternal reinforcement. The learning process is divided into several stages, with each stage designed to ensure mastery before progressing to the next. Students are rewarded with grades or other forms of positive reinforcement when they successfully learn the material. This reinforcement stimulates further learning and motivation. Unlike the Western focus on external behaviors, the Russian school emphasizes managing internal (mental) processes. This includes fostering logical and rational methods and ways of thinking. The emphasis is on the development of mental actions through structured guidance and scaffolding. The goal is to internalize cognitive processes so that students can independently manage their learning. Fundamental instruction methodologies are centred on the formation students' ability to think logically and rationally. In order to achieve this, thoroughly structured assignments which develop the student's cognitive abilities gradually are used.
Adaptive learning is a comprehensive approach to education that combines methodological principles, instructional processes, and technological tools to personalize learning and optimize student outcomes. By leveraging data-driven insights and dynamic adaptation, adaptive learning empowers educators to meet the diverse needs of students and foster deeper learning experiences. P. Kerr's considers adaptive learning to be an educational technology aimed at creating "automated, dynamic and interactive" content [4]. Lowen-dahl J.M. determines adaptive learning as a process. His definition of adaptive learning as a process aligns with the broader understanding of adaptive learning as a dynamic and iterative approach to education. [5]. N. Yalaeva and her colleagues consider adaptive English language teaching as a "method that uses technology to customize and individualize the learning process for each student. This method provides the opportunity to develop an individualized learning program that incorporates the level of knowledge, interests and needs of each learner [14]". In our opinion, adaptive learning represents a multifaceted approach that integrates technology, method, and process to enhance learning outcomes. By leveraging adaptive web applications and systems, educators can provide personalized feedback and support that caters to individual students' needs and learning styles. This holistic approach ensures that adaptive learning remains responsive to the dynamic nature of education and supports students in achieving their full potential.
The research conducted through various scholarly databases like Web of Science, Scopus, Cyberleninka, and Elibrary highlights several key themes regarding the design of adaptive learning systems. Researchers emphasize the importance of considering various student characteristics to create effective and personalized adaptive learning experiences. These characteristics include: basic knowledge, learning styles, cognitive styles or thinking styles, metacognitive knowledge, student preferences, student's abilities, student behavior, student profile (gender, age, sex, etc.), and student's interests.
Indeed, alongside empirical studies, literature reviews play a crucial role in advancing our understanding of adaptive learning technologies and their various features. Scholars offer comprehensive reviews on different aspects of adaptive learning, including: intellectual and cognitive abilities of students, adaptive educational hypermedia, characteristics of Students, adaptive learning systems, and adaptive technologies.
At present, although there are many reviews on adaptive learning, there is still no comprehensive study available that would trace the dynamics of adaptive learning development, explore current approaches to its practical application, examine global educational practices, and outline future directions. Moreover, most of the existing reviews focus on technical applications, neglecting the deep historical roots of adaptive learning ideas. Therefore, our goal is to fill this gap by conducting a literature review that examines in detail the historical development of adaptive learning.
Methodology
As part of our research, we used a comprehensive literature review that is supposed to show the dynamics of adaptive learning development.Through this method, we systematically reviewed, evaluated, and summarised the existing literature on the evolution of adaptive learning from its inception to its current practical application. Our goal was to provide a detailed description of how adaptive learning has evolved over time, highlighting major milestones, influential theories, and technological advances.
We began by defining the scope of our literature review, focusing on the historical progression of adaptive learning. Our objectives were to document significant developments, identify major contributors, and understand the evolution of adaptive learning theories and technologies.
Analysis of the results
The roots of adaptive learning can be traced back to ancient Greece, specifically to Socrates (469-399 BCE). Socrates employed a method of teaching that was highly interactive and personalized. Known as the Socratic Method, this approach involved asking a series of probing questions to stimulate critical thinking and illuminate ideas. This form of dialogue was adaptive in nature, as Socrates would tailor his questions based on the responses and understanding of his students, guiding them towards deeper insight and knowledge.
During the Renaissance and Enlightenment periods, educational theories began to emphasize the importance of individualized learning. Educators like John Amos Comenius (1592-1670) advocated for teaching methods that considered the individual needs and abilities of students. Comenius's ideas laid the groundwork for future educational reforms that would seek to adapt instruction to the learner.
The following stage in the development of adaptive learning is the 1950s-1960s. John Dewey, an American philosopher and educator, was a leading propo-
C3
o
CO "O
1=1 A
—I
o
C3 t; o m o
OT
3
u o
CO
o d
u
OJ
o
CM
nent of progressive education, which emphasized experiential learning and the active role of the student in the learning process. Dewey's work laid the foundation for modern adaptive teaching by promoting the idea that education should be tailored to the needs and interests of each student.
Lev Vygotsky (1896-1934), a Russian psychologist, introduced the concept of the Zone of Proximal Development (ZPD), which describes the difference between what a learner can do independently and what he/she can achieve with guidance. Vygotsky's theories highlighted the importance of providing adaptive support (scaffolding) to help students progress within their ZPD.
B.F. Skinner (1904-1990), an American behavio-rist, pioneered programmed instruction, an early form of adaptive teaching. While effective in shaping student behavior through reinforcement, Skinner's approach didn't account for students' prior knowledge and individual differences. Nonetheless, his work paved the way for advancements in adaptive learning and personalized instruction. [41].
N. Crowder (1921-1998) elaborated a complex algorithm of programmed learning to adjust learning in accordance with students' answers. Using multiple-choice tasks, he assessed the degree of students' understanding of the material and determined the subsequent training stage. Nevertheless, due to its complexity, this extended learning algorithm was not widely used for educational purposes.
Another influential figure in the development of adaptive learning algorithms is G. Pask (1928-1996). He devised an algorithm that differentiated tasks based on varying levels of difficulty to accommodate different students' abilities. This approach aimed to tailor the learning experience to each student's needs, ensuring appropriate challenge and support for optimal learning outcomes.
Both Russian scholars and Western researchers have contributed to the study of adaptive and programmed learning.
L.N. Landa (1927-1999) was a pioneering Russian scholar who introduced the concept of an algorithm of mental actions, revolutionizing psychology's understanding of learning processes. Landa's work laid the foundation for the application of algorithmization in education, shaping modern approaches to instructional design and personalized learning.
A.S. Granitskaya's adaptive teaching method prioritizes personalization, differentiation, and flexibility in instruction to meet the diverse needs of individual learners. It involves continuous assessment, varied instructional strategies, and individualized support. While offering benefits such as increased engagement and improved learning outcomes, successful implementation requires careful planning and ongoing professional development.
According to Galperin, learning occurs through the internalization and mastery of external activities or operations. This theory emphasizes the importance of guiding learners through structured stages of activity, from external regulation to internalized control, to fa-
cilitate meaningful learning. Galperin's approach has had a profound impact on educational practices, particularly in shaping instructional methods that scaffold learning experiences and promote the development of higher-order thinking skills. His theory outlines a six-stage process: 1) motivational foundation of action; 2) teacher's explanation; 3) external action (students perform the action); 4) verbal action (practicing the action through speech); 5) mental action (internal speech); 6) performing the action mentally.
Benjamin Bloom (1913-1999) was a renowned American educational psychologist best known for his taxonomy of educational objectives, commonly referred to as Bloom's Taxonomy. This framework categorizes educational goals into cognitive domains, ranging from lower-order thinking skills, such as remembering and understanding, to higher-order skills, such as analyzing, evaluating, and creating. Bloom's Taxonomy has been widely used in curriculum design, instructional planning, and assessment development, providing educators with a structured framework to promote critical thinking and intellectual development in learners. Additionally, Bloom made significant contributions to the field of mastery learning, advocating for personalized instruction and competency-based assessment to ensure that all students achieve mastery of essential concepts and skills before progressing to more advanced material. His work continues to influence educational practices worldwide, shaping pedagogical approaches that prioritize the development of higher-order thinking skills and the cultivation of lifelong learners.
The 1970s and 1980s saw a significant intersection between cognitive theories and technological advancements. Cognitive theories, particularly in psychology and neuroscience, were gaining traction during this time period, influencing various fields including education, artificial intelligence, and human-computer interaction. These theories emphasized the importance of understanding mental processes such as perception, memory, attention, and problem-solving.
Technological advancements, particularly in computing and information technology, were also rapidly progressing during this era. Researchers and engineers began applying insights from cognitive theories to develop innovative technologies aimed at enhancing human cognition, improving learning, and creating more intuitive human-computer interfaces.
In the early 1990s, adaptive hypermedia systems emerged, blending adaptive interfaces and user model-based interfaces within hypermedia environments. P. Brusilovsky and his team were pioneers in this field, focusing on developing adaptive hypermedia systems to offer personalized navigation support for users on web pages. These systems utilize user models to tailor information and links according to each individual's needs [1].
The integration of AI and machine learning in education has significantly advanced adaptive teaching. Modern adaptive learning platforms, such as Duolin-go, and Smart Sparrow, use sophisticated algorithms to analyze student data and personalize learning paths
in real-time. These systems can adapt content, pace, and instructional strategies based on continuous assessment of student performance.
The use of big data and learning analytics has provided educators with detailed insights into student learning behaviors and outcomes. This information allows for more precise adaptation of teaching methods and materials to meet individual student needs.
Adaptive learning has a rich history marked by challenges in implementation, especially in comparison to strategies devised by expert educators. Modern models of adaptive learning, though, have made significant progress in overcoming these challenges. Three basic models are typically identified: the student model, the content model, and the instructional model.
The student model serves as the cornerstone, providing vital information about individual students necessary for an adaptive system. It includes a variety of data such as personal information, academic performance, topics and tests taken, video views, reading supplemental materials, and grades. With this information, the system predicts the learner's behaviour and tailors the learning process to individual needs. Adapting learning materials based on a learner model is particularly popular because it is believed that the effectiveness of an adaptive system depends on its ability to adjust to different learner characteristics.
The instructional model, also known as the pedagogical model, plays a crucial role in adapting learning experiences based on the learner model. It determines decisions about the content to be delivered to the student, including aspects such as pace, format, and sequence. At the same time, the content model focuses on the characteristics of the content itself, finding appropriate learning resources that meet the needs and preferences of the users.These models are based on various algorithms and data analysis methods. Modern research explores multiple approaches to implementing adaptive learning. One such approach involves designing the educational process with consideration for different student characteristics. For instance, Markovic S. has developed a system that customizes learning experiences based on individual student profiles. [7].
Adaptive learning is recognized as a fundamental approach for accommodating individual differences by broadening the spectrum of teaching methods. With the increasing popularity and advancement of technologies, alongside supportive software and hardware for adaptive learning, UNESCO asserts that it is widely accepted that all learners should benefit from adaptive learning initiatives (UNESCO Institute for Information Technologies in Education, 2020).
While adaptive learning is a promising concept, there is still a notable lack of data on its effectiveness. Current research suggests that its efficacy should be compared to the traditional teacher-centered approach, where the instructor plays the central role. Adaptive learning's main method, which focuses on the needs and abilities of students, is the student-centered approach.
However, C. Perrotta and B. Williamson point out that in some cases, the broad implementation of adaptive learning might not reduce educational inequality. Instead, it could widen the gap between students of different socioeconomic backgrounds. Additionally, many companies developing adaptive educational programs are often motivated by profit rather than educational principles" [9].
As "new computer technologies are introduced into the educational process, a unified educational and cultural space is formed, the professional competence of teachers is improved" [13] adaptive learning is gaining popularity, drawing interest from educators, psychologists, business professionals, and administrators. With the increasing importance of higher education, universities face the challenge of offering a personalized approach to ensure the most effective learning for each student.
The anticipated benefits of adaptive learning technology are significant: increased student motivation, engagement, better educational outcomes, improved job market readiness, and lower dropout rates. However, the COVID-19 pandemic has profoundly impacted higher education. Starting in early 2020, students worldwide transitioned to online learning, moving traditional lectures and seminars to a new, unfamiliar format. Although online technologies quickly became mainstream, their quality and effectiveness are still in question.
The lack of agreement on the effectiveness of adaptive educational systems in higher education may result from both limited empirical data and inherent challenges in adaptive learning itself. These challenges include data collection for both the subject matter and student models, affecting the entire adaptive algorithm cycle. Specific problems include a limited range of disciplines for adaptation (mostly technical disciplines), difficulties in measuring complex constructs, significant financial and time investments, and limited adaptation capabilities.
Conclusion
Our comprehensive literature review reveals that adaptive teaching has a long and rich history, from its roots in the Socratic method of ancient Greece to the sophisticated digital platforms of today. Over time, adaptive learning has evolved through significant milestones and technological advancements, continually improving its ability to cater to individual learning needs.
In contemporary education, adaptive learning is gaining popularity, promising increased student motivation, engagement, better educational outcomes, improved job market readiness, and lower dropout rates. The COVID-19 pandemic has accelerated the shift to online learning, highlighting both the potential and challenges of adaptive learning technologies.
Despite challenges such as data collection, complexity in measuring learning constructs, and significant investment requirements, the future of adaptive learning remains highly promising. As technology advances, adaptive learning systems are poised to rev-
C3
o
CO "O
1=1 A
—I
o
C3 t; o
m o
OT
3
u o
CO
olutionize education by providing personalized experiences that effectively meet the diverse needs of all
students.
Литература
1. Brusilovsky P. Methods and techniques of adaptive hypermedia // Adaptive hypertext and hypermedia. 1998. P. 1-43.
2. Carbonell J. R. AI in CAI: An artificial-intelligence approach to computer-assisted instruction //IEEE transactions on man-machine systems. - 1970. -Т. 11. - № . 4. - С. 190-202.
3. Crowder N.A. Automatic tutoring by means of intrinsic programming //Automatic teaching: The state of the art. - 1959. - Т. 116.
4. Kerr P. Adaptive learning // ETL Journal. 2016. Vol. 70. № 1. P. 89.
5. Lowendahl J.M., Thayer T.L.B., Morgan G. Top 10 strategic technologies impacting higher education in 2016 [Электронный ресурс] // Gartner Research. 2016. URL: https://www.academia. edu/29441505/Top_10_Strategic_Technologies_ Impacting_Higher_Education_in (дата обращения: 17.05.2024)
6. Lowendahl J. M. et al. Top 10 strategic technologies impacting higher education in 2016 //Research Note G. - 2016. - Т. 294732. - С. 15.
7. MarkoviC S. MODELING OF ADAPTIVE SYSTEM FOR DISTANCE LEARNING WITH AN EMPHASIS ON STUDENT PROFILE //Journal of Education Research. - 2015. - Т. 9. - № . 3.
8. Pask G. Electronic keyboard teaching machines // Education and Commerce. - 1958. - Т. 24. -С.16-26.
9. Perrotta C., Williamson B. The social life of Learning Analytics: cluster analysis and the 'performance' of algorithmic education // Learning, Media and Technology. 2018. Т. 43. № . 1. P. 3-16.
10. Skinner B.F. Teaching Machines: From the experimental study of learning come devices which arrange optimal conditions for self-instruction //Science. - 1958. - Т. 128. - № . 3330. - С. 969-977.
11. Wenger E. Artificial Intelligence and Education. -1987.
12. Xu D., Wang H., Su K. Intelligent student profiling with fuzzy models //Proceedings of the 35th Annual Hawaii international conference on system sciences. - IEEE, 2002. - С. 8 pp.
13. Yalaeva, N.V. The main principles of developing students' language competence in English classes at a law school / N.V. Yalaeva, N.V. Sadykova // Modern Pedagogical Education. - 2021. - No. 7. -P. 108. - EDN VZPPXZ.
14. Ялаева, Н.В. Применение технологий адаптивного компьютерного обучения при подготовке студентов-юристов / Н.В. Ялаева, Н.В. Сады-кова, Т.В. Кудинова // Современное педагогическое образование. - 2023. - № 9. - С. 227. -EDN SGGMTZ.
EVOLUTION OF ADAPTIVE LEARNING METHODS: ASPECTS OF INTRODUCING ARTIFICIAL INTELLIGENCE TECHNOLOGIES
Kudinova T.V.
MIREA Russian Technological University
The article deals with the evolution of adaptive learning, tracing its origins to ancient Greece, and the personalized interactive Socrat-ic method. This historical approach laid the foundation for modern adaptive learning practices. Over the centuries, significant milestones in educational theory, set by John Amos Comenius, Lev Vy-gotsky, and Burrhus Frederic Skinner, have gradually shaped adaptive learning. The author stresses that technological advances in the late 20th century, particularly in computer-assisted learning and cognitive tutoring, have further developed adaptive learning methods. The integration of artificial intelligence and machine learning into modern adaptive learning platforms has revolutionized education by enabling real-time personalized learning. Despite such challenges as data collection, complexity, and the need for significant investment, the future of adaptive learning remains promising. It has the potential to increase student motivation and engagement, improve learning outcomes, enhance employability, and reduce dropout rates. However, challenges remain, including difficulties in adapting to a wide range of disciplines, measuring complex constructs, and the time and financial costs required for effective implementation.
Keywords: adaptive learning, individualized learning, programmed learning, educational technology, AI, machine learning.
References
1. Brusilovsky P. Methods and techniques of adaptive hypermedia // Adaptive hypertext and hypermedia. 1998. P. 1-43.
2. Carbonell J. R. AI in CAI: An artificial-intelligence approach to computer-assisted instruction //IEEE transactions on man-machine systems. - 1970. - Т. 11. - № . 4. - С. 190-202.
3. Crowder N.A. Automatic tutoring by means of intrinsic programming //Automatic teaching: The state of the art. - 1959. - Т. 116.
4. Kerr P. Adaptive learning // ETL Journal. 2016. Vol. 70. № 1. P. 88-93. DOI:10.1093/elt/ccv055
5. Lowendahl J.M., Thayer T.L.B., Morgan G. Top 10 strategic technologies impacting higher education in 2016 [Электронный ресурс] // Gartner Research. 2016. URL: https://www.academ-ia.edu/29441505/Top_10_Strategic_Technologies_Impacting_ Higher_Education_in (дата обращения: 17.05.2024)
6. Lowendahl J. M. et al. Top 10 strategic technologies impacting higher education in 2016 //Research Note G. - 2016. -Т. 294732. - С. 15.
7. Markovic S. MODELING OF ADAPTIVE SYSTEM FOR DISTANCE LEARNING WITH AN EMPHASIS ON STUDENT PROFILE //Journal of Education Research. - 2015. - Т. 9. - № . 3.
8. Pask G. Electronic keyboard teaching machines //Education and Commerce. - 1958. - Т. 24. - С. 16-26.
9. Perrotta C., Williamson B. The social life of Learning Analytics: cluster analysis and the 'performance' of algorithmic education // Learning, Media and Technology. 2018. Т. 43. № . 1. P. 3-16.
10. Skinner B.F. Teaching Machines: From the experimental study of learning come devices which arrange optimal conditions for self-instruction //Science. - 1958. - Т. 128. - № . 3330. -С. 969-977.
11. Wenger E. Artificial Intelligence and Education. - 1987.
12. Xu D., Wang H., Su K. Intelligent student profiling with fuzzy models //Proceedings of the 35th Annual Hawaii international conference on system sciences. - IEEE, 2002. - С. 8 pp.
13. Yalaeva, N.V. Application of adaptive computer learning technologies in law students training / N.V. Yalaeva, N.V. Sadykova, T.V. Kudinova // Modern Pedagogical Education. - 2023. -№ 9. - С. 226-231. - EDN SGGMTZ.
14. Yalaeva, N.V. The main principles of developing students' language competence in English classes at a law school / N.V. Yalaeva, N.V. Sadykova // Modern Pedagogical Education. - 2021. - No. 7. - P. 108-111. - EDN VZPPXZ.
