Research Landscape of Adaptive Learning in Education: A Bibliometric Study on Research Publications from 2000 to 2022

Abstract

Adaptive learning is an approach toward personalized learning and places the concept of “learner-centered education” into practice. With the rapid development of artificial intelligence and other technologies in recent years, there have been many breakthroughs in adaptive learning. Thus, it is important to gain insight into the evolution of related research and to track the research frontiers to further promote its development. This study used CiteSpace and VOSviewer to conduct a bibliometric analysis of 644 adaptive learning journal papers indexed in the WoS database from 2000 to 2022. This study presented a general view of the field of adaptive learning research over the last two decades using quantitative analysis. Currently, adaptive learning research is rapidly developing. In terms of the major research forces, a core group of authors including Qiao J. F., Han H. G. and Song Q has been formed; the major publishing country in this field is China; the core publishing journals include IEEE Transactions on Neural Networks and Learning Systems. Four major research topics in this field were identified using cluster analysis, namely the application of deep learning in educational data analysis, the development and application of adaptive learning model in AI education, the development and application of intelligent tutoring system in tutoring and teaching, cutting-edge modeling technology for feature modeling and knowledge tracing. Through evolution analyses, the logic of adaptive learning research’s development was determined; that is, technological changes have played a key role in the development of this field. Following the logic, we presented three frontiers of adaptive learning with burst terms: feature extraction, adaptation model and computational modeling. Adaptive learning is a core research topic for both computer science and educational technology disciplines, and it is also an important field where emerging technologies empowering education and teaching can play a part. The findings of the study clearly presented the current research status, evolutionary logic and research frontiers of this topic, which can provide references for the further development of this research field.

1. Introduction

Adaptive learning refers to a learning method that dynamically adjusts to the goal or type of course content based on an individual’s cognition and abilities, and it provides personalized support and boosts the learner’s performance using both technology and instructor interventions [1]. The technology that supports the realization of such a method is called adaptive learning technologies. In recent years, the prevalence of smart devices such as mobile terminals and the rapid development of emerging technologies, including artificial intelligence, cloud computing, big data and adaptive learning technologies, boosted the research and practice of adaptive learning [2], a field that has also become a trending topic and research focus in education [3]. Meanwhile, adaptive learning technologies are considered promising in the field of interdisciplinary research [4]. For example, in the 2018 National Survey of eLearning and Information Technology in U.S. Higher Education, Green (2018) found that most leaders with a higher education background have a positive attitude towards adaptive learning and think that it has great potential in helping students achieve success [5]. In addition, the EDUCAUSE Horizon Report 2020 included adaptive learning technologies in the six key technologies that will influence the future of higher education [6].

Over the past two decades, adaptive learning has received attention from computer science, artificial intelligence and educational research, and many scholars have creatively applied advanced algorithms in artificial intelligence to education, which not only yielded a series of research results in the field [7,8,9,10,11], but also prompted many breakthroughs [12]. It can be argued that research on adaptive learning has borne some fruit. However, several questions remain unknown: What are the current major research focuses in adaptive learning? What is the status quo and trend of relevant research? What is the current research frontier or trend? In addition, there have been few in-depth, systematic visual analyses of the research results in this field. This has caused problems: First, it hinders the clarification of the overview of current research on adaptive learning and the further improvement of the entire research system. Second, it prevents researchers from tracking current research progression and from grasping key and trending research topics, based upon which researchers can precisely pinpoint and examine subsequent research developments. Given the above situation, the present study intends to draw upon Hwang and Tu’s [13] study and use bibliometric methods to conduct a systematic visual analysis of the existing research studies on adaptive learning between 2000 and 2022 from three perspectives—major research forces (core authors, publishing countries and publishing journals), the status quo of research (topic clusters and key research points), research trend and frontier—to tease out and present the development, research progress, research focus and trends in this field. Therefore, this study centers on the following three research questions.

• (1). Between 2000 and 2022, in the field of adaptive learning, who were the major authors? What were the top publishing countries? What were the major journals that included papers on adaptive learning research?

• (2). What were the main research topics (research clusters) in adaptive learning between 2000 and 2022, and what points did these topics specifically focus on?

• (3). What was the development trend of adaptive learning between 2000 and 2022? What is the current trending topic for research in the field?

2. Literature Review of Adaptive Learning

Adaptive learning is an important means that accommodates individual differences via the diversification of teaching methods [1,14]. With the popularization and maturement of technology and the software and hardware that support adaptive learning, it is generally acknowledged as a policy that all students shall benefit from adaptive learning (UNESCO Institute for Information Technologies in Education, 2020) [15].

The first research on adaptive learning dates back to the world’s first adaptive test—the Binet IQ test by the French psychologist Binet, introduced in 1905 [16]. An in-depth analysis of the research findings reveals significant differences between what was previously and currently referred to as adaptive learning. Specifically, in the past, adaptive learning was more about the teaching pattern supported by teaching-centered automatic teaching machines, such as the first automatic teaching machine designed by Pressey [17], an American psychologist, in the 1920s, which provided students with multiple-choice questions and tracked learning responses. The machine is also regarded as the beginning of computer-assisted instruction. Later, as technology evolved, the technology supporting adaptive learning was upgraded from early traditional teaching intervention machines (e.g., IQ test) to intelligent information platforms. For example, in 1996, the world’s first real adaptive learning system, AEHS (Adaptive Educational Hypermedia System), was born. It comprised a combination of a hypermedia/hypertext system, an adaptive system and an intelligent teaching system, marking the gradual maturement of adaptive learning systems [14]. Recent years have witnessed the emergence of AI-combined adaptive products and their related applications with the rise of artificial intelligence technology. Adaptive platforms represented by Knewton, Smart Sparrow and Cerego can offer learners personalized learning materials and timely learning feedback, transforming learning content from traditional uniform delivery to on-demand personalized delivery, focusing extensively on learners’ needs and enhancing their learning experience.

With the advent of various adaptive learning platforms, the theoretical and empirical research studies related to adaptive learning have become more diversified. For example, some scholars have verified the important influence of learners’ cognitive and learning styles on their choice of adaptive learning environments [18], and they have proposed that the theoretical framework of adaptive learning environments should include three modules: the learning content development module, the adaptable presentation module and the adaptive content module.

As technology is essential to the development of adaptive learning, the relationship between technology and adaptive learning has also been the center of attention for scholars. Some scholars have pointed out the need to pay attention to the issue of technological rationality when pursuing the effective use of technology, and they have suggested that the purpose of technology is to facilitate education rather than lead it [19]. With the academic field’s deepening understanding of adaptive learning, research perspectives are becoming more diverse. For example, Peng et al. [7] proposed that the four core elements of adaptive learning are personal characteristics, personal performance, personal development and adaptive adjustment. Regarding technology-enhanced adaptive learning, Xie et al. [8] suggested that there is still significant room for improvement concerning the application of adaptive learning to smart devices. Bernacki et al. [20] summarized the theoretical guidance method of personalized adaptive learning via a literature survey and employed it to evaluate learning practices.

In summary, supported by emerging technologies, adaptive learning rapidly developed and became a focal research topic in education worldwide [3,21]. Empirical research has been conducted by learners to investigate technologies that support adaptive learning and the pedagogical theories that underpin them—mostly around algorithm-based resource recommendations and system-based instructional interventions. However, no in-depth systematic visualization of the results available has been found by the authors. The few existing relevant studies are those that have analyzed technology-enhanced personalized learning and adaptive learning between 2007 and 2017 by using systematic literature reviews [8]. However, the number of self-adaptive learning studies has rapidly grown in the past five years. To assist relevant researchers and practitioners in terms of providing a more comprehensive and timely overview of the latest research on adaptive learning, such as major research forces, current research status, research trends and research hotspots, and to promote the sound development of subsequent research studies, it is necessary to comprehensively track and conduct a visual analysis of relevant research by starting from the stage when adaptive learning systematically developed (2000). With this purpose, this study adopts a bibliometric approach; in terms of time span, it systematically analyzes research studies related to the field of adaptive learning between 2000 and 2022; in terms of research content, it focuses on research forces by analyzing core authors, publishing journals and countries, and on the research status by analyzing research themes and key points; then, this study systematically analyzes an overview of adaptive learning research in combination with evolutionary trends and research trends.

3. Methodology and Materials

3.1. Bibliometric and Research Tools

Bibliometrics is a metrology method that combines mathematics, statistics and bibliography. It quantitatively analyzes and classifies a comprehensive body of knowledge on research [22] by studying abstracts of the existing literature [23]. In 1969, Pritchard first proposed that statistical bibliography should be replaced with bibliometrics [24]. Today, bibliometrics has become the center of the internal logical structure of information science, as it has the advantage of conducting a quantitative and intuitive analysis of the literature using knowledge mapping and cluster analysis [25]. It can be argued that bibliometrics supports the analysis of the development of a particular field from multidimensional perspectives, such as sorting out the current state of research in the field, tracking research trends, and presenting major research forces [26]. Among them, sorting out the current status of research can help identify the shortcomings and the focus of existing research studies; tracking research trends helps grasp the direction of subsequent research studies [27]; presenting major research forces helps clarify the main researchers and institutions by using various statistical indicators (e.g., h-index and number of citations).

This study mainly adopted a bibliometric method by using VOSviewer (Version: 1.6.18, developed by Van Eck and Waltman, Leiden, The Netherlands) and CiteSpace (Version: 6.1.R2, developed by Chen C., Philadelphia, PA, USA) to map the core papers on adaptive learning in the WoS database between 2000 and 2022. VOSviewer was used to conduct statistical analyses on countries and authors, analyze the major research forces in adaptive learning and carry out co-occurrence and cluster analyses with respect to keywords [28]. As for CiteSpace, it was employed to perform evolution analyses, measure the similarity of data units via a set-theoretic approach relative to data standardization, and accordingly generate a time-zone view, which chronologically presents the process of the evolution of research on adaptive learning and the change in research trends [29]. The software was also used to identify the current trends in adaptive learning via burst terms detection, which provides an overview of the field’s development, as well as its future trends.

3.2. Data Retrieval

The first step of data retrieval in a bibliometric study is to choose the databases and specific indexes used in the study. The present study selected the Web of Science (WoS), Science Citation Index-Expanded (SCI-Expanded), and Social Science Citation Index (SSCI) databases as the data source for the following reasons.

• (1). WoS is currently considered to be the most suitable database for bibliometric analyses [30], and the database has been frequently used for econometric analyses in research areas such as electronic information, economy and educational research [31,32,33].

• (2). WoS constitutes the most reliable and high-quality citation database in the world [34,35]. The SCIE and SSCI in this database are the most authoritative indexes in the field of natural and social sciences [36]. The literature included in the two major indexes has undergone strict double-blind peer review processes and has significantly higher inclusion criteria than that of the Scopus database. Therefore, to ensure the quality of the literature included in the analysis, the two major indexes in the WoS database were used as the data source for the analysis in this study.

The time span of this study was set from January 2000 to August 2022, with the specific end time being 15 August. The reason for choosing this time span is that, although the concept of adaptive learning was created early, it has not been put into practice for a long period of time due to immature technology and the cost of application. With the emergence of online learning in the late 20th and early 21st century, adaptive learning became better externally supported, and systematic research on adaptive learning began to appear at the beginning of this century [12,37]. Therefore, this study selected the literature published from 2000 to 2022 as the data source.

A summary of the preliminary data retrieval is shown in Table 1, and a total of 1610 relevant papers were obtained.

3.3. Data Filtering

Data accessed from databases without careful screening can be problematic due to the duplication of literature, inconformity with the research theme or substandard completeness and other problems. Therefore, it is necessary to perform data screening and standardization before the analysis in order to avoid affecting the analysis’s accuracy due to the quality of the data. The following three main steps were performed to check and screen the data.

• (1). Firstly, we address the duplication of the literature. This was performed by the “remove duplicated” function of CiteSpace, and the selection criteria were based on the DOI of all papers. After the analysis, two duplicates were found, and 1608 samples remained after the deletion.

• (2). Secondly, we deal with the inconformity of research themes. To solve the problem, team members had discussions with experts in adaptive learning to set the selection criteria (see Table 2). It is important to observe that adaptive learning is also a computing terminology, and this study only focuses on adaptive learning techniques used in educational research, so computer science research only focusing on adaptive learning techniques was not included. Based on such criteria, the literature was independently screened by three team members, and controversial instances were discussed to determine if it met the requirements. Finally, 644 papers that did not match the theme were excluded, and 964 papers were retrieved.

• (3). Finally, we standardize the data. Nguyen and Hallinger [38] pointed out that metadata exported from scientific databases often contain multiple expressions of the same term, which, if not rationalized in bibliometric analysis, would lead to repetition in the keyword co-occurrence network. For example, in our metadata set, the list of keywords included “neural network”, “neural networks (nns)”, “neural-network” and “neural-networks”. To forestall any adverse impact on the study introduced by problems within the data itself, a data disambiguation process was carried out and applied to the metadata set before data analyses [39]. This step mainly relied on VOSviewer and manual merging. The txt data files downloaded from the database were uploaded to VOSviewer, and problematic data terms were identified and recorded. Then, manual data cleaning was performed following the criteria that when words are of the same meaning, low-frequency words alternate to high-frequency ones, and nouns take priority when the frequencies were similar.

  The flow of data retrieval and data filtering is presented in Figure 1.

4. Performance Analysis

4.1. The Development Trend of Research on Adaptive Learning Research

The number of publications is an important indicator of the development of a research field. Moreover, the change in annual publication numbers, in particular, not only indicates the prosperity of the field to a certain extent, but also visibly presents the macrotrends. Figure 2 shows the annual publication volume in adaptive learning research from 2000 to 2022. Taken together, the number of publications in the period shows different features before and after 2017: In the period from 2000 to 2017, the annual number of publications was stable, mostly fluctuating around 30. The publication volume shows a significant growth trend starting from 2018, with the annual publication number exceeding 60 in the years between 2018 and 2022, and up to 123 in 2021. This suggests that adaptive learning research is receiving increasing attention from scholars and is flourishing.

4.2. Quantitative Analysis of the Authors

An analysis of the author’s publication volume provides information regarding the representative scholars and core research strengths of a research area. Table 3 presents information on the core authors in adaptive learning research, including names, publication number and the average number of citations per article.

The research focus of different scholars considerably varies. Qiao J. F. is the most prolific author in the field of adaptive learning research in the last 20 years, with a total of 13 publications. His main research areas are artificial-intelligence-driven automation [40,41], intelligent computing and intelligent optimal control [42,43]. In addition, he closely worked with scholars such as Han H. G. [40,41,44]. Vamvoudakis K. G. is a well-known scholar in adaptive learning and has published seven papers related to adaptive learning research in the past two decades, with an average of 105.29 citations per paper. His research focuses are interdisciplinary in nature, including control theory [45], reinforcement learning [46,47] and cyber-physical systems [48].

4.3. Quantitative Analysis of the Countries

To learn about which countries have made the most significant contributions to adaptive learning research, the present study analyzed the publication number of sixty-eight countries, among which the top ten are presented in Table 4. An analysis of the data in Table 4 shows that China is the country with the highest number of publications in adaptive learning research, with 437 publications (10,190 citations), accounting for 45% of the total publications in the field. The U.S. and India follow China with 153 (5875 citations) and 52 (857 citations) publications, respectively. Moreover, an analysis of the geographic location shows that the top ten countries are all in Asia, North America and Europe.

4.4. Quantitative Analysis of the Journals

Journals are the main vehicle for publishing relevant research results. Therefore, the present study analyzed the top ten journals publishing most articles on adaptive learning and calculated the average number of citations per article (Table 5).

As observed in Table 5, the journal with the most publications in the field of adaptive learning research is IEEE Transactions on Neural Networks and Learning Systems (58 articles), followed by IEEE Access and Neurocomputing, which reflects the fact that current adaptive learning research is often associated with deep learning analytics, such as neural networks and neurocomputing. In addition, many journals with a high number of publications belong to the field of educational technology, such as Computer Assisted Language Learning; ERT&D—Educational Technology Research and Development; Computer Applications in Engineering Education; and Educational Technology and Society, which, to some extent, represent the interdisciplinary nature of this research topic; this confirms that adaptive learning is an exemplar of evolving computer technologies applied to education and a typical educational application empowered by technology that meets educational demands.

A further review of the journals including papers on adaptive learning reveals that there are no journals closely related to the topic of adaptive learning yet, suggesting that most of the current research on adaptive learning depends on the development of related technological disciplines (e.g., computing disciplines) or educational technology.

5. Co-Occurrence Analysis and Evolution Analysis

5.1. Co-Occurrence Analysis on Keywords

Keywords usually serve as the focus of a study, and it is widely accepted that high-frequency keywords in some certain fields can somewhat represent research highlights. Based on the co-occurrence relationship of high-frequency keywords, the overall development trend of one research field can be learned via cluster analyses. Wei et al. [49] proposed that Price’s Law can not only analyze the core authors in a field but also determine high-frequency keywords. In accordance with formula M = 0.749 ∗ $\sqrt{{\mathrm{N}}_{\max }}$—in which N_max stands for the keyword with the highest frequency of occurrence and, according to the VOSviewer, it is known to be 136—M is approximately equal to 8.73. Therefore, keywords with a frequency higher than 8 are taken as the high-frequency keywords in the field. This study uses VOSviewer to analyze the co-occurrence relationship of high-frequency keywords in 964 papers in order to determine this field’s research highlights in different periods and their level of concern [50]. The visualization atlas of the results is shown in Figure 3.

It can be learned from Figure 3 that the field of adaptive learning from 2000 to 2022 mainly focused on four topics, namely ”the application of deep learning in education data analysis” (red clustering), “the development and application of adaptive learning models in AI education” (green clustering), “the development and application of intelligent guidance systems in learning guidance and teaching” (blue clustering) and “the cutting-edge modeling technology serving feature modeling and knowledge tracking” (yellow clustering). In addition, it can be observed in the figure that the clusters of “the application of deep learning in education data analysis” and “the development and application of adaptive learning models in AI education” are closely related, while the clusters of “the development and application of intelligent guidance systems in learning guidance and teaching” and “the cutting-edge modeling technology serving feature modeling and knowledge tracking” are relatively independent. In addition, a few keywords simultaneously appeared in the two clusters, such as “neural network” and other keywords in the adaptive learning model cluster and stability optimization control cluster, which shows that these two clusters obtain both their own meanings and some common characteristics of cross research.

Red clustering refers to “the application of deep learning in education data analysis”. The main keywords include natural network, algorithm, network, optimization, adaptive learning rate, conversion, deep learning, classification, identification, etc. This clustering method aims to explore the modeling and data analysis of educational data supported by deep learning, which is represented by neural networks [51,52,53]. The deep learning (DL) algorithm is applied to many aspects, with educational data mining and learning analysis as the major ones. Analysis technologies, which include AI and data mining with unique computing, prediction and interaction functions, have attracted increasing attention due to their potential for supporting teaching and learning [54]. For example, the collaborative optimization algorithm [55] uses the stochastic properties of ant colony optimization and the exploratory properties of the genetic algorithm for an optimal solution, providing learners with personalized learning methods. In addition, the full path learning recommendation model [56], based on the similarity of their features, clusters learners to train their long short-term memory (LSTM). With the optimal learning path as a foundation, it can offer reasonable recommendations. This method significantly improved the accuracy of results and the efficiency of learning. In the research of Dwivedi et al. [57], the learning path recommendation system tends to focus on online learners’ needs, preferences and their level of knowledge, and it recommends optimal learning paths for individuals via variable length genetic algorithms, which are based on collaborative filtering. In addition, this technology also plays a unique and vital role in the field of special education. Based on iSocial, a three-dimensional collaborative game for social ability development—a verbal and nonverbal interaction test driven by virtual images of autistic learners—shows that social theory and cluster analyses appear to be effective in identifying unique patterns of verbal and nonverbal interactions involving different levels of concrete social existence [58]. Specifically, the effectiveness of various forms of educational equipment and their applications, such as Chatterbot [59,60], digital games [61,62,63] and computer-mediated communication [64], mainly benefits from effective integrated analysis technologies such as data mining [65,66] and machine learning [67]. These technologies enable us in conducting a profound study on regular education patterns in massive data and building relevant prediction models. Therefore, better decision-making can be carried out.

Green clustering is “the development and application of adaptive learning models in AI education”, with the main keywords being model, performance, design, systems, strategies, framework, etc. This clustering method aims to study the development and application of adaptive intelligent learning systems or models [68,69,70,71]. Earlier researchers usually focused on the development of learning resources based on SCORM standards and student learning models. For instance, the object-oriented curriculum model created by Tseng et al. [72] arranges teaching materials into different types and transforms them into modular learning objects, enabling the thematic learning content to be dynamically formed in accordance with each student’s need; in addition, Yaghmaie and Bahreininejad built an adaptive learning management system framework based on the multi-agent system and learning process to improve learning quality [73]. With the diversification of analysis methods, more scholars have tended to integrate methods, such as cognitive diagnosis [74,75] and visual diagnosis [76,77], to analyze teaching models in order to better tap into students’ learning potential. In addition, scholars have studied and explained students’ online learning behavior with an adaptive learning cognitive map model, accordingly providing online learning suggestions. For example, the adaptive learning cognitive map model proposed by Wan and Yu [78] and the online learning system based on it can provide learners with adaptive learning resources and opportunities for self-reflection. In recent years, as the subject group of adaptive learning research has expanded, learners in pre-service education and elementary education have been included in the research study. Combining games with adaptive learning is now a key approach for satisfying personalized learning demands and conducting adaptive learning for young learners [79]. A general evaluation of the research included in the clustering obtained one commonality: adaptive learning has the unique potential to examine the specific learning targets, priori knowledge, cognitive ability and the background of learners.

Blue clustering is “the development and application of intelligent guidance systems in learning guidance and teaching”, and the main keywords are nonlinear-system, adaptive learning system, uncertainty, tracking, adaptive control, stability, iterative learning control, etc. This clustering method is aimed to assist technologies such as nonlinear systems and neural networks [80], as well as their specific methods when applied to learning tutorship [81,82,83] and machine control demonstration teaching [84]. With the development of new AI technologies such as neural network and big data, the educational value of adaptive tutoring system gradually became prominent, and the relevant educational applications started to attract increasing attention from scholars. Many used adaptive learning technology to diagnose, evaluate and guide learners [85,86]. Caballé et al. [87], for instance, proposed a new virtual learning resource for collaborative learning, known as collaborative complex learning resources (CC-LR), which enables students in carrying out effective adaptive learning by means of interaction and collaboration. In addition, based on the concept–effect relationship, Chu et al. [88] developed a diagnosis and remedy system to detect students’ learning problems, which not only enhances students’ performance, but also improves their attitude and self-efficacy. Furthermore, the development and popularization of technologies outside the scope of AI have also facilitated the application and promotion of adaptive tutoring systems. Due to the fact that these technologies hardly rely on neural networks or other deep learning technologies for which their processing and principles are difficult to explain, they became more promotional and can better serve precision teaching. Luo [76], for instance, used eye-tracking technology to identify learning methods and proved that this technology could quickly identify different types of learners via empirical research. However, there are also interferences that may lead to different levels in terms of recognition accuracy, such as the small sample size of the study, and these interferences still need to be removed or improved in future research.

Yellow clustering is a cutting-edge modeling technology serving feature modeling and knowledge tracking, and it contains some main keywords such as adaptation model, task analysis, optimal control, tracking control, information learning, etc. The clustering method targets the method used in modeling and tracking learners’ recent or long-term learning progression [89,90,91]. Narciss et al. [92], for example, tracked students’ progress and activities with ActiveMath, a web-based intelligent learning environment, to present score tasks. In addition, Sharma, Papamitsiou and Giannakos [93] highlighted the importance of multimodal data in the adaptive modeling process, and they improved machine learning algorithms to analyze multimodal data, thus realizing the evaluation and prediction of learners’ performance. Meanwhile, this study is also committed to solving the inexplicability of deep learning; that is, by combining the “white-box” approach driven by the hypothesis/literature (feature extraction) with the “black-box” approach driven by computation/data (feature fusion), the interpretation of input data and output results can be realized to a certain extent. In addition, the advanced optimization algorithm of deep learning is also eye-catching in this cluster [94,95]. For example, Wang and Liu [96] conducted in-depth research on the association rule mining algorithm. By improving the FP-Tree Algorithm of the algorithm itself and data sources, the learner feature model of modern music classrooms is established. In addition, the COVID-19 pandemic emphasized the importance of virtual learning environments (VLE). With machine learning, students can more effectively track and process a multitude of data generated by daily interactions with VLE. The construction of an adaptive learning system under this clustering method also becomes more sophisticated. The adaptive learning system designed and developed by Adnan et al. [97], for example, enables more accurate recommendations of learning resources by assigning different weights to different input data, such as a learner’s performance data, which provides a basis for teachers or administrators to optimize learning content while also providing students with more targeted learning guidance to improve learners’ learning experience.

In short, using clustering analyses, it can be concluded that AI technology is of great significance relative to adaptive learning, and new AI technologies enable the emergence of high-quality research. Based on keyword clustering, we can figure out the core technology and common application scenarios in the field of adaptive learning. There is no doubt that the continuous development and integration of emerging technologies further expanded the breadth and depth of adaptive learning application scenarios, gradually becoming an important force for the advancement of this field.

5.2. Evolution Analysis and Research Frontiers

Based on evolutionary logic and the latest frontiers, the development trends of the adaptive learning research can be learned from the dual perspectives of development history and frontier topics. Therefore, this study analyzes the evolution of high-frequency keywords in the time zone of the selected literature by using CiteSpace to track and observe the dynamic development process and frontier topics throughout the research development of adaptive learning (Figure 4). Moreover, this study systematically presents changes in the research frontiers of adaptive learning by using the burst detection of CiteSpace (Table 6).

From the distribution and the density of keywords, it can be observed that the development of adaptive learning in the past 20 years has shown significantly phased characteristics, and these can be divided into three stages: The period from 2000 to 2007, when a large number of basic keywords of this field emerged, can be regarded as the first stage. From 2008 to 2013, although there were few high-frequency keywords, many technical words were poured into this research field; thus, this period can be seen as the second stage. From 2014 to 2022, with the development of new technologies, new terms emerged, many of which were maintained as the focuses of current research (Table 6); thus, this period can be regarded as the third stage.

5.2.1. Evolution Analysis of the First Stage (2000–2007)

It can be observed from the review of the research in the first stage that the concepts of adaptive learning and adaptive learning systems were relatively mature at the beginning of the 21st century. On the basis of the emergence and application of e-learning systems, scholars developed adaptive learning systems, and they attempted to self-adjust and reorganize learning materials and paths according to learners’ feedback to cater to learner’s needs of interests, competence and behavior [98]. However, research in the first stage was mainly based on the principles of pedagogy when designing adaptive learning systems. For example, Wang [99] designed an instruction guidance system with adaptive function by virtue of scaffolding based on Vygotsky’s “zone of proximal development” theory. Some scholars studied the important issues in adaptive learning research, such as resource selection, ranking, knowledge tracking, etc. [100]. However, on the one hand, due to the shallow cognition of self-adaptation in this stage, technology and education were not yet deeply integrated; on the other hand, although some research achievements accurately grasped some valuable research questions and proposed creative ideas, their results failed to comprehensively guide the practice of education and in a timely manner, because of the limitations of software and hardware and the insufficient ability of technology and information.

In addition, in this stage, some scholars started to apply AI technology in adaptive learning. Keywords in the AI field, such as algorithm and neural network, appeared at the beginning of this stage and have appeared throughout the development of adaptive learning for more than 20 years. However, the application of AI technology was, at the same time, mainly focused on machine learning, including fuzzy logic, Bayesian networks, etc. [101,102]. Across the early studies in the first stage, it can be observed that most research studies at that time relied on the analysis of learners’ past learning records. This method provided an opportunity for assessing learners’ cognitive load, but there were also some prominent problems: For example, equipment was necessary to track learners’ study situations for a long period of time. Therefore, this method had a defect in terms of real-time evaluation, and it was difficult to realize large-scale popularization and application.

5.2.2. Evolution Analysis of the Second Stage (2008–2013)

In the second stage, research on adaptive learning presents specifically diversified developments in technology and factors. First of all, the learning factors focused on by scholars became comprehensive. For example, more personalized factors (such as cognitive style) pertaining to learners were included in research [18,103] in order to achieve precise technical empowerment, which is shown in the emergence of keywords, such as learning style, in Figure 4. In addition, new technology was expanding, and more AI-related technologies, such as decision trees [104], expert systems [105] and particle swarm optimization [7,106], were introduced into this research field. Later in this stage, the popularity of personal computers (e-learning) and mobile devices (mobile learning) greatly promoted the development of adaptive learning research and also provided an environment for its promotion and application [71].

By comparing Figure 4 and Table 6, it can be observed that there are two closely related explosive words in this stage: particle swarm optimization and genetic algorithm. Particle swarm optimization is an evolutionary algorithm proposed at the end of the 20th century, which is also a key algorithm that boosted the development of AI [107]. Based on the observation of the group and activity behavior, the algorithm can make use of the individual’s information shared within a group to generate an evolutionary process of the group’s movement from disordered to ordered in the problem-solving process, and thus obtain the optimal solution. After more than ten years of development, with the maturity of particle swarm optimization and its great application value in adaptive learning, it is now a cutting-edge technology, providing the optimal solution for recommended resources [106]. The genetic algorithm has also received attention for a long period of time. Compared with particle swarm optimization, it has more underlying meanings. However, similar to particle swarm optimization, genetic algorithms are also methods for seeking the optimal solution by simulating the natural evolution process. The two methods intersect in concepts but are not coincident, and both are applied in adaptive learning research, providing the optimal solution for the recommendation of adaptive learning resources. Up until now, related studies based on the two algorithms were still important parts of adaptive learning research [108].

5.2.3. Evolution Analysis and Frontier of the Third Stage (2013–2022)

In the third stage, adaptive learning research presents more detailed technical factors, such as deep learning, identification, tracking control, etc., among which the most typical is deep learning (DL), which is a new direction of machine learning. The purpose of introducing DL into machine learning is to enable machine learning to realize self-adaptation and ultimately achieve intelligent adaptive learning [109]. Specifically, DL works by learning the inherent laws of sample data and has the ability to interpret data, such as words, images, sounds, etc. Its ultimate goal is to enable machines to recognize, analyze and study in a similar manner to that of human beings [108]. The large-scale promotion of DL is key to the realization and development of adaptive learning, because the technology for intelligently perceiving and evaluating learners’ cognitive load and knowledge level is fundamental for the realization of adaptive learning. Zhang et al. [110] pointed out that DL technology opens up a new path for adaptive learning research, and is an important technical direction that is worthy of focus in the third stage of adaptive learning research.

However, from Table 6, it can be observed that although DL was introduced into the research of adaptive learning in 2014, it has not been widely applied in adaptive learning, and it did not become a focus of research in this field until 2019. This phenomenon reflects the diachronic relationship between the emergence, development and application of DL technology, and it also shows that the transformation and optimization of technology based on research needs are necessary in order to integrate technologies into research and to be functional in educational practices. After the technology with DL was introduced into the research of adaptive learning as core methods, it introduced revolutionary influences with respect to the development of adaptive learning, and it is now at the forefront of current research studies. As observed in Table 6, in addition to DL, other core technologies at the forefront of adaptive learning research include feature extraction, adaptation models and computational modeling.

Feature extraction is a proper term in DL, and it refers to the process of transforming raw data into a better representation of the potential characteristics of prediction objects. Thanks to the inclusiveness of DL relative to data (DL is able to accept poorly structured, multi-formed, multi-modal data), DL can expand the collection scope of learners’ personalized information, including non-standard data such as eye movements and facial expressions, for application in analyses. Moreover, feature extraction is the process of extracting the information required for training models from the data. With the support of feature extraction and the verification of multi-modal and multi-dimensional data, increasingly accurate measurements of learners’ cognitive and development levels can be realized [93].

The adaptation model is an adaptive system with dynamic adjustment functions supported by DL technology. Based on the acquisition of learners’ data, the system has the ability to adjust and update itself in real time. However, many problems still remain in the current research state of the adaptation model, such as the contradiction between real-time computing and the needed amount of time for collecting and analyzing large amounts of learners’ data [111].

In addition, the development of computational modeling based on DL makes it possible for adaptive learning to build a high-precision adaptive model that tracks learners’ development in real time. It also provides a solution for the problems existing in the current research of the adaptation model and supports the realization of adaptive learning [112].

6. Conclusions, Implications and Future Research

6.1. Conclusions

This research study used bibliometrics to track international adaptive learning research trends based on the core literature in adaptive learning research over the past 20 years in the WoS database; the following conclusions were obtained.

• (1). Research on adaptive learning is now developing at high speeds. China is currently the main publishing country in this field, contributing more than 40% of research papers. Moreover, Chinese scholars make up the majority of the core authors in this field, with core authors represented by Qiao J. F., Han H. G. and Song Q., who have made great contributions to the development of adaptive learning. Currently, there have been a number of core journals in adaptive learning research that focus on the fields of computer science (e.g., IEEE Transactions on Neural Networks and Learning Systems and IEEE Access) and education technology (e.g., Computer Assisted Language Learning and ERT&D—Educational Technology Research and Development) (RQ1).

• (2). For the cluster analysis of the relevant literature in adaptive learning research, there are four focused research topics, namely the following: “the application of deep learning in education data analysis”, “the development and application of adaptive learning models in AI education”, “the development and application of intelligent guidance systems in learning guidance and teaching” and “the cutting-edge modeling technology serving feature modeling and knowledge tracking”. These four clusters all have similar technical bases: various analysis technologies and algorithms for machine learning and deep learning in the category of artificial intelligence. Based upon these focal topics, the evolution of technology has received the most attention, followed by the design, development and application of the system and the construction of a model of the characteristics of learners (RQ2).

• (3). The evolution analysis and research frontier present the development history of adaptive learning in the past 20 years and the current research frontier. The main trend in its development is that the upgrades with respect to technology and applications boost the rapid development of adaptive learning research. The current research frontier focuses on the DL, which is still changing and developing, and the frontier includes feature extraction, adaption models and computational modelling; these are the three frontiers that receive the most amount of attention from scholars. Feature extraction focuses on the standardized processing of various data, while adaptation models and computational modeling pay more attention to building real-time and dynamic models (RQ3).

6.2. Findings and Implications

Through bibliometric methods and the screening of the core literature in the research field based on the results of the quantitative analysis, the present study systematically analyzed the research overview of adaptive learning of the past two decades. The study also identified the four major research themes in the field based on cluster analysis, the evolutionary logic of the field based on keyword time zones and the current research frontiers based on word-burst analysis.

In addition, the research study also reveals that social change, technology and educational needs are the key factors driving the development of adaptive learning, and that with the empowerment of emerging technologies, the ethical issue in terms of technology in adaptive learning is beginning to attract attention. This suggests that we should, for one thing, capture the exogenous variables that promote adaptive learning, such as social and technological changes, in order to inject energy and momentum into adaptive learning; for another, we should fully grasp the endogenous variables that promote adaptive learning based on educational needs and consider the demands and directions of adaptive learning development from the perspective of educational challenges and bottlenecks so as to iteratively empower educational development and form a two-way approach between changes and demands. Third, we should pay attention to the limits of technological applications, reasonably use technology to empower adaptive learning research and application and avoid undesirable behaviors, including technology-led education and the invasion of personal privacy, so that adaptive learning can truly be used to maximize its value in helping to achieve personalized learning in a healthy and effective way.

The theoretical and practical implications of the above findings include the following: Firstly, it provides an innovative and systematic overview of adaptive learning research over the past 20 years in terms of research forces, research themes, evolutionary trends and frontiers, which enriches theoretical research on adaptive learning and helps researchers and practitioners in establishing a clearer perception of its value. Secondly, the visualization of research forces can help researchers and practitioners understand the core authors, core journals and core publishing countries in the field, providing guidance and references for the subsequent construction of research communities and increase literature reviews and paper submission. Third, via an analysis of the theme and evolutionary logic of adaptive learning research, the main contents receiving attention in the field and their evolution in different periods are clearly presented to researchers and practitioners, thus helping researchers judge the research trends in adaptive learning and providing references for scholars when enriching existing theories and selecting research topics. Finally, analysis of the frontiers can help researchers and practitioners with the identification of the current focus of the research field, and provide references for subsequent research.

6.3. Limitations and Future Research

Adaptive learning, as a highly active and cutting-edge cross topic with profound theoretical foundations, undoubtedly has great development potential. It is reasonable to believe that, with the development of technology, research on adaptive learning will have broad prospects in this era of revolution. Therefore, this research area is significant, and its value should be further enhanced.

Influenced by objectivity, this study also has some limitations. Firstly, bibliometrics software imparts high standards on data resources. In order to ensure the quality and integrity of the collected data, only journal papers in SSCI and the SCI-Expanded Index in the Web of Science Core Collection were selected for this research study, and some research studies on adaptive learning may have been ignored as a result of this. Secondly, due to space limitations, there are still some areas that this research study did not cover. For example, this study did not systematically summarize the technology commonly used in current adaptive learning research; it did not deeply analyze the advantages and disadvantages of current technologies in adaptive learning; it did not discuss the ethical problems of the application of new technology in current adaptive learning research in education and teaching. From another perspective, these limitations also point out directions for future research. Future studies can combine more scientific methods, such as systematic literature reviews, to obtain more detailed information in this field, including but not limited to the application of specific technologies, the type and scale of collected data, etc. It is also suggested that meta-analyses technology can be further utilized to accurately assess the value of the application of adaptive learning systems.

Author Contributions

Conceptualization, Y.J. and L.Z.; methodology, C.W. and H.W.; software, L.Z.; validation, Y.J., L.Z. and K.Z.; formal analysis, L.Z. and H.W.; data curation, L.Z.; writing—original draft preparation, Y.J., C.W. and L.Z.; writing—review and editing, K.Z. and Q.X.; visualization, L.Z. and Q.X.; supervision, Q.X.; project administration, Y.J.; funding acquisition, Y.J. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Acknowledgments

The authors would like to express their sincere gratitude to the reviewers and editor for their valuable suggestions, to College of Educational Science and Technology, Zhejiang University of Technology for the cultivation.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Data retrieval, filtering and standardization process (PRISMA figure).

Figure 2. Time trend of the publications on adaptive learning.

Figure 3. Co-occurrence of keywords.

Figure 4. Keyword evolution analysis.

Table 1

Summary of data source and selection.

Category	Specific Standard Requirements
Research database	Web of Science core collection
Citation indexes	SCI-Expanded and SSCI
Searching period	January 2000 to August 2022
Language	“English”
Searching keywords	“adaptive learning” OR “adaption model” OR “adaptive learning system” OR “adaptive system”
Subject categories	“Education Scientific Disciplines” OR “Computer Science Artificial Intelligence” OR “Computer Science Information Systems” OR “Automation Control Systems” OR “Education Educational Research” OR “Computer Science Theory Methods” OR “Computer Science Interdisciplinary Applications” OR “Computer Science Software Engineering” OR “Psychology Multidisciplinary”
Document types	“Articles”
Sample size	1610

Table 2

Inclusion and exclusion criteria.

Inclusion Criteria	Exclusion Criteria
The research topic of the paper focuses on adaptive learning, such as theoretical studies on adaptive learning, the construction of adaptive learning models and studies related to the application of adaptive technologies for promotion.	The research topic of the paper is not relevant to adaptive learning. For example, it only mentions adaptive learning as the research background; adaptive-learning-related content accounts for a small proportion (less than 10%) of the research
The paper contains at least three pages	The paper is a report numbered less than three pages, a short essay or an introduction
Full-text papers are available	The full text is unavailable because of various reasons (e.g., retraction)
The paper is fully equipped with necessary information (e.g., abstract, author’s information, keywords and references)	The paper is seriously lacking in necessary information and is hard to complete (e.g., abstract, author’s information, keywords and references)
The paper includes research questions, methods and conclusions (results)	The paper does not clearly present research questions, methods and conclusions

Table 3

Most important authors in the adaptive learning research field.

Rank	Author	Documents	Citations	Average Citation per Paper
1	Qiao, J. F.	13	312	24
2	Han, H. G.	9	253	28.11
3	Song, Q.	9	210	23.33
4	Ganjefar, S.	8	106	13.25
5	Vamvoudakis, K. G.	7	737	105.29
6	Lin, F. J.	7	554	79.14
7	Mu, C. X.	7	350	50
8	Tomei, P.	7	169	24.14
9	Cai, J. P.	7	40	5.71
10	Yan, Q. Z.	7	40	5.71
11	He, H. B.	6	552	92
12	Hou, Z. S.	6	233	38.83

Table 4

Top 10 countries in the adaptive learning research field.

Rank	Country	Documents	Citations	Average Citation per Paper
1	China	437	10190	29.37
2	The United States	153	5875	38.28
3	India	52	857	16.48
4	Iran	46	1288	28
5	England	43	1477	34.35
6	South Korea	35	431	12.31
7	Spain	35	393	11.23
8	Singapore	34	786	23.12
9	Canada	33	873	26.45
10	Italy	29	419	14.45

Table 5

Top 10 journals in the adaptive learning research field.

Rank	Source	Documents	IF
1	IEEE Transactions on Neural Networks and Learning Systems	58	14.255
2	IEEE Access	49	3.476
3	Neurocomputing	46	5.779
4	Expert Systems with Applications	32	8.665
5	Computer Assisted Language Learning	18	5.964
6	ERT&D—Educational Technology Research and Development	17	5.580
7	Computers in Human Behavior	17	8.957
8	Computer Applications in Engineering Education	16	2.109
9	Educational Technology and Society	15	2.633
10	International Journal of Control, Automation and Systems	14	2.964

Table 6

Burst keywords in adaptive learning research from 2000 to 2022.

Rank	Keywords	Strength	Begin	End	Time Distribution
1	network	4.42	2000	2006	▃▃▃▃▃▃▃ ▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂
2	learning control	3.62	2007	2012	▂▂▂▂▂▂▂ ▃▃▃▃▃▃ ▂▂▂▂▂▂▂▂▂▂
3	particle swarm optimization	4.32	2013	2016	▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃▃▃▃ ▂▂▂▂▂▂
4	genetic algorithm	3.88	2013	2015	▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃▃▃ ▂▂▂▂▂▂▂
5	design	3.87	2015	2018	▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃▃▃▃ ▂▂▂▂
6	identification	4.64	2014	2019	▂▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃▃▃▃▃▃ ▂▂▂
7	deep learning	5.23	2019	2022	▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃▃▃▃
8	feature extraction	3.59	2019	2022	▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃▃▃▃
9	adaptation model	9.93	2020	2022	▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃▃▃
10	computational modeling	4.22	2020	2022	▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃▃▃

References

1. Capuano, N.; Caballé, S. Adaptive learning technologies. AI Mag.; 2020; 2, pp. 96-98. [DOI: https://dx.doi.org/10.1609/aimag.v41i2.5317]

2. Peng, H.; Ma, S.; Spector, J.M. Personalized adaptive learning: An emerging pedagogical approach enabled by a smart learning environment. Smart Learn. Environ.; 2019; 1, pp. 1-14. [DOI: https://dx.doi.org/10.1186/s40561-019-0089-y]

3. Wang, S.; Christensen, C.; Cui, W.; Tong, R.; Yarnall, L.; Shear, L.; Feng, M. When adaptive learning is effective learning: Comparison of an adaptive learning system to teacher-led instruction. Intract. Learn. Environ.; 2020; pp. 1-11. [DOI: https://dx.doi.org/10.1080/10494820.2020.1808794]

4. Cavanagh, T.; Chen, B.; Lahcen, R.A.M.; Paradiso, J.R. Constructing a design framework and pedagogical approach for adaptive learning in higher education: A practitioner’s perspective. Int. Rev. Res. Open. Dis.; 2020; 1, pp. 173-197. [DOI: https://dx.doi.org/10.19173/irrodl.v21i1.4557]

5. Campus Computing 2018: The 29th National Survey of Computing and Information Technology in American Higher Education. Available online: https://www.campuscomputing.net/content/2018/10/31/the-2018-campus-computing-survey (accessed on 26 December 2022).

6. 2022 EDUCAUSE Horizon Report|Teaching and Learning Edition. Available online: https://library.educause.edu/-/media/files/library/2022/4/2022hrteachinglearning.pdf (accessed on 26 December 2022).

7. Peng, T.; Zuo, W.; He, F. SVM based adaptive learning method for text classification from positive and unlabeled documents. Knowl. Inf. Syst.; 2008; 3, pp. 281-301. [DOI: https://dx.doi.org/10.1007/s10115-007-0107-1]

8. Xie, H.; Chu, H.C.; Hwang, G.J.; Wang, C.C. Trends and development in technology-enhanced adaptive/personalized learning: A systematic review of journal publications from 2007 to 2017. Comput. Educ.; 2019; 140, 103599. [DOI: https://dx.doi.org/10.1016/j.compedu.2019.103599]

9. Martin, F.; Chen, Y.; Moore, R.L.; Westine, C.D. Systematic review of adaptive learning research designs, context, strategies, and technologies from 2009 to 2018. ETRD-Educ. Technol. Res.; 2020; 4, pp. 1903-1929. [DOI: https://dx.doi.org/10.1007/s11423-020-09793-2]

10. Villegas-Ch, W.; Roman-Cañizares, M.; Jaramillo-Alcázar, A.; Palacios-Pacheco, X. Data analysis as a tool for the application of adaptive learning in a university environment. Appl. Sci.; 2020; 20, 7016. [DOI: https://dx.doi.org/10.3390/app10207016]

11. Normadhi, N.B.A.; Shuib, L.; Nasir, H.N.M.; Bimba, A.; Idris, N.; Balakrishnan, V. Identification of personal traits in adaptive learning environment: Systematic literature review. Comput. Educ.; 2019; 130, pp. 168-190. [DOI: https://dx.doi.org/10.1016/j.compedu.2018.11.005]

12. Li, F.; He, Y.; Xue, Q. Progress, challenges and countermeasures of adaptive learning. Educ. Technol. Soc.; 2021; 24, pp. 238-255.

13. Hwang, G.J.; Tu, Y.F. Roles and research trends of artificial intelligence in mathematics education: A bibliometric mapping analysis and systematic review. Mathematics; 2021; 6, 584. [DOI: https://dx.doi.org/10.3390/math9060584]

14. Brusilovsky, P. Methods and techniques of adaptive hypermedia. Adaptive Hypertext and Hypermedia; Springer: Dordrecht, The Netherlands, 1998; Volume 1, pp. 1-43.

15. UNESCO. International Forum on AI and the Futures of Education, Developing Competencies for the AI Era, 7–8 December 2020: Synthesis Report Synthesis Report; UNESCO: Paris, France, 2020.

16. Binet, A.; Simon, T. New methods for the diagnosis of the intellectual level of subnormals. (L’Année Psych., 1905, pp. 191–244). The Development of Intelligence in Children (The Binet-Simon Scale); Binet, A.; Simon, T.; Kite, E.S. Williams & Wilkins Co.: Philadelphia, PA, USA, 1916; pp. 37-90.

17. Pressey, S.L. A machine for automatic teaching of drill material. Sch. Soc.; 1927; 25, pp. 549-552.

18. Yang, T.C.; Hwang, G.J.; Yang, S.J.H. Development of an adaptive learning system with multiple perspectives based on students’ learning styles and cognitive styles. Educ. Technol. Soc.; 2013; 4, pp. 185-200.

19. Sarıyalçınkaya, A.D.; Karal, H.; Altinay, F.; Altinay, Z. Reflections on Adaptive Learning Analytics: Adaptive Learning Analytics. Advancing the Power of Learning Analytics and Big Data in Education; IGI Global: Hershey, PA, USA, 2021; Volume 1, pp. 61-84.

20. Bernacki, M.L.; Greene, M.J.; Lobczowski, N.G. A systematic review of research on personalized learning: Personalized by whom, to what, how, and for what purpose (s)?. Educ. Psychol. Rev.; 2021; 4, pp. 1675-1715. [DOI: https://dx.doi.org/10.1007/s10648-021-09615-8]

21. Mo, C.Y.; Wang, C.; Dai, J.; Jin, P. Video Playback Speed Influence on Learning Effect from the Perspective of Personalized Adaptive Learning: A Study Based on Cognitive Load Theory. Front. Psychol.; 2022; 13, 839982. [DOI: https://dx.doi.org/10.3389/fpsyg.2022.839982]

22. Diodato, V.P.; Gellatly, P. Dictionary of Bibliometrics; 1st ed. Routledge: New York, NY, USA, 2013; [DOI: https://dx.doi.org/10.4324/9780203714133]

23. Donthu, N.; Kumar, S.; Pattnaik, D. Forty-five years of Journal of Business Research: A bibliometric analysis. J. Indian Bus. Res.; 2020; 109, pp. 1-14. [DOI: https://dx.doi.org/10.1016/j.jbusres.2019.10.039]

24. Pritchard, A. Statistical bibliography or bibliometrics?. J. Doc.; 1969; 4, pp. 348-349.

25. Verbeek, A.; Debackere, K.; Luwel, M.; Zimmermann, E. Measuring progress and evolution in science and technology-I: The multiple uses of bibliometric indicators. Int. J. Manag. Rev.; 2002; 2, pp. 179-211. [DOI: https://dx.doi.org/10.1111/1468-2370.00083]

26. Yang, Y.; Wang, C.; Lai, M. Using bibliometric analysis to explore research trend of electronic word-of-mouth from 1999 to 2011. Int. J. Innov. Technol.; 2012; 3, pp. 337-342. [DOI: https://dx.doi.org/10.7763/IJIMT.2012.V3.250]

27. Wang, C.L.; Dai, J.; Xu, L.J. Big data and data mining in education: A bibliometrics study from 2010 to 2022. In Proceeding of the 2022 7th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA); Chengdu, China, 22–24 April 2022; [DOI: https://dx.doi.org/10.1109/ICCCBDA55098.2022.9778874]

28. Van Eck, N.J.; Waltman, L. Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics; 2010; 2, pp. 523-538. [DOI: https://dx.doi.org/10.1007/s11192-009-0146-3]

29. Chen, J.; Wang, C.; Tang, Y. Knowledge mapping of volunteer motivation: A bibliometric analysis and cross-cultural comparative study. Front. Psychol.; 2022; 13, 883150. [DOI: https://dx.doi.org/10.3389/fpsyg.2022.883150] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35719581]

30. Ding, X.; Yang, Z. Knowledge mapping of platform research: A visual analysis using VOSviewer and CiteSpace. Electron. Commer. Res.; 2020; 22, pp. 1-23. [DOI: https://dx.doi.org/10.1007/s10660-020-09410-7]

31. Popescu, D.V.; Dima, A.; Radu, E.; Dobrotă, E.M.; Dumitrache, V.M. Bibliometric Analysis of the Green Deal Policies in the Food Chain. Amfiteatru. Econ.; 2022; 24, pp. 410-428. [DOI: https://dx.doi.org/10.24818/EA/2022/60/410]

32. Dima, A.; Bugheanu, A.M.; Boghian, R.; Madsen, D.Ø. Mapping Knowledge Area Analysis in E-Learning Systems Based on Cloud Computing. Electronics; 2023; 12, 62. [DOI: https://dx.doi.org/10.3390/electronics12010062]

33. Chen, J.Y.; Liu, Y.D.; Dai, J.; Wang, C.L. Development and status of moral education research: Visual analysis based on knowledge graph. Front. Psychol.; 2023; 13, 1079955. [DOI: https://dx.doi.org/10.3389/fpsyg.2022.1079955]

34. Birkle, C.; Pendlebury, D.A.; Schnell, J.; Adams, J. Web of Science as a data source for research on scientific and scholarly activity. Quant. Sci. Stud.; 2020; 1, pp. 363-376. [DOI: https://dx.doi.org/10.1162/qss_a_00018]

35. Cobo, M.J.; Martínez, M.A.; Gutiérrez-Salcedo, M.; Fujita, H.; Herrera-Viedma, E. 25 years at knowledge-based systems: A bibliometric analysis. Knowl.-Based Syst.; 2015; 80, pp. 3-13. [DOI: https://dx.doi.org/10.1016/j.knosys.2014.12.035]

36. Pranckute, R. Web of Science (WoS) and Scopus: The Titans of Bibliographic Information in Today’s Academic World. Publications; 2021; 9, 12. [DOI: https://dx.doi.org/10.3390/publications9010012]

37. Fernandez, G. Cognitive scaffolding for a web-based adaptive learning environment. International Conference on Web-Based Learning (ICWL); Springer: Berlin/Heidelberg, Germany, 2003; [DOI: https://dx.doi.org/10.1007/978-3-540-45200-3_2]

38. Nguyen, U.P.; Hallinger, P. Assessing the distinctive contributions of simulation & gaming to the literature, 1970–2019: A bibliometric review. Simulat. Gaming; 2020; 6, pp. 744-769. [DOI: https://dx.doi.org/10.1177/1046878120941569]

39. Strotmann, A.; Zhao, D. Author name disambiguation: What difference does it make in author-based citation analysis?. J. Am. Soc. Inf. Sci. Technol.; 2012; 9, pp. 1820-1833. [DOI: https://dx.doi.org/10.1002/asi.22695]

40. Han, H.; Wu, X.; Qiao, J. Nonlinear systems modeling based on self-organizing fuzzy-neural-network with adaptive computation algorithm. IEEE Trans. Cybern.; 2014; 4, pp. 554-564. [DOI: https://dx.doi.org/10.1109/TCYB.2013.2260537]

41. Qiao, J.; Li, F.; Han, H.; Li, W. Growing Echo-State Network with Multiple Subreservoirs. IEEE. Trans. Neur. Net. Learn.; 2017; 2, pp. 391-404. [DOI: https://dx.doi.org/10.1109/TNNLS.2016.2514275]

42. Gu, K.; Zhou, J.; Qiao, J.F.; Zhai, G.; Lin, W.; Bovik, A.C. No-Reference Quality Assessment of Screen Content Pictures. IEEE. T. Image. Process.; 2017; 8, pp. 4005-4018. [DOI: https://dx.doi.org/10.1109/TIP.2017.2711279]

43. Gu, K.; Tao, D.; Qiao, J.F.; Lin, W. Learning a No-Reference Quality Assessment Model of Enhanced Images with Big Data. IEEE. Trans. Neur. Net. Learn.; 2018; 4, pp. 1301-1313. [DOI: https://dx.doi.org/10.1109/TNNLS.2017.2649101]

44. Han, H.G.; Chen, Q.L.; Qiao, J.F. An efficient self-organizing RBF neural network for water quality prediction. Neural. Netw.; 2011; 7, pp. 717-725. [DOI: https://dx.doi.org/10.1016/j.neunet.2011.04.006]

45. Vamvoudakis, K.G.; Lewis, F.L. Online actor–critic algorithm to solve the continuous-time infinite horizon optimal control problem. Automatica.; 2010; 5, pp. 878-888. [DOI: https://dx.doi.org/10.1016/j.automatica.2010.02.018]

46. Lewis, F.L.; Vrabie, D.; Vamvoudakis, K.G. Reinforcement learning and feedback control: Using natural decision methods to design optimal adaptive controllers. IEEE. Contr. Syst. Mag.; 2012; 6, pp. 76-105. [DOI: https://dx.doi.org/10.1109/mcs.2012.2214134]

47. Kiumarsi, B.; Vamvoudakis, K.G.; Modares, H.; Lewis, F.L. Optimal and autonomous control using reinforcement learning: A survey. IEEE Trans. Neur. Net. Learn.; 2017; 6, pp. 2042-2062. [DOI: https://dx.doi.org/10.1109/TNNLS.2017.2773458]

48. Bhasin, S.; Kamalapurkar, R.; Johnson, M.; Vamvoudakis, K.G.; Lewis, F.L.; Dixon, W.E. A novel actor–critic–identifier architecture for approximate optimal control of uncertain nonlinear systems. Automatica; 2013; 1, pp. 82-92. [DOI: https://dx.doi.org/10.1016/j.automatica.2012.09.019]

49. Wei, W.; Ge, J.; Xu, S.; Li, M.; Zhao, Z.; Li, X.; Zheng, J. Knowledge maps of disaster medicine in China based on co-word analysis. Disaster Med. Public; 2019; 3, pp. 405-409. [DOI: https://dx.doi.org/10.1017/dmp.2018.63]

50. Bodily, R.; Leary, H.; West, R.E. Research trends in instructional design and technology journals. Br. J. Educ. Technol.; 2019; 1, pp. 64-79. [DOI: https://dx.doi.org/10.1111/bjet.12712]

51. De Melo, F.R.; Flôres, E.L.; De Carvalho, S.D.; De Teixeira, R.A.G.; Loja, L.F.B.; de Sousa Gomide, R. Computational organization of didactic contents for personalized virtual learning environments. Comput. Educ.; 2014; 79, pp. 126-137. [DOI: https://dx.doi.org/10.1016/j.compedu.2014.07.012]

52. Holmes, M.; Latham, A.; Crockett, K.; O’Shea, J.D. Near real-time comprehension classification with artificial neural networks: Decoding e-learner non-verbal behavior. IEEE Trans. Learn. Technol.; 2017; 1, pp. 5-12. [DOI: https://dx.doi.org/10.1109/TLT.2017.2754497]

53. Dutt, S.; Ahuja, N.J.; Kumar, M. An intelligent tutoring system architecture based on fuzzy neural network (FNN) for special education of learning disabled learners. Educ. Inf. Technol.; 2022; 2, pp. 2613-2633. [DOI: https://dx.doi.org/10.1007/s10639-021-10713-x]

54. Chen, X.; Zou, D.; Xie, H.; Chen, G.; Lin, J.; Cheng, G. Exploring contributors, collaborations, and research topics in educational technology: A joint analysis of mainstream conferences. Educ. Inf. Technol.; 2022; pp. 1-36. [DOI: https://dx.doi.org/10.1007/s10639-022-11209-y]

55. Vanitha, V.; Krishnan, P.; Elakkiya, R. Collaborative optimization algorithm for learning path construction in E-learning. Comput. Electr. Eng.; 2019; 77, pp. 325-338. [DOI: https://dx.doi.org/10.1016/j.compeleceng.2019.06.016]

56. Zhou, Y.; Huang, C.; Hu, Q.; Zhu, J.; Tang, Y. Personalized learning full-path recommendation model based on LSTM neural networks. Inf. Sci.; 2018; 444, pp. 135-152. [DOI: https://dx.doi.org/10.1016/j.ins.2018.02.053]

57. Dwivedi, P.; Kant, V.; Bharadwaj, K.K. Learning path recommendation based on modifed variable length genetic algorithm. Educ. Inf. Technol.; 2018; 2, pp. 819-836. [DOI: https://dx.doi.org/10.1007/s10639-017-9637-7]

58. Wang, X.; Xing, W.; Lafey, J.M. Autistic youth in 3D game-based collaborative virtual learning: Associating avatar interaction patterns with embodied social presence. Brit. J. Educ. Technol.; 2018; 4, pp. 742-760. [DOI: https://dx.doi.org/10.1111/bjet.12646]

59. Chen, H.; Park, H.W.; Breazeal, C. Teaching and learning with children: Impact of reciprocal peer learning with a social robot on children’s learning and emotive engagement. Comput. Educ.; 2020; 150, 103836. [DOI: https://dx.doi.org/10.1016/j.compedu.2020.103836]

60. Zhang, H.; Huang, T.; Liu, S.; Yin, H.; Li, J.; Yang, H.; Xia, Y. A learning style classification approach based on deep belief network for large-scale online education. J. Cloud. Comput.; 2020; 1, pp. 1-17. [DOI: https://dx.doi.org/10.1186/s13677-020-00165-y]

61. Bang, H.J.; Li, L.; Flynn, K. Efficacy of an Adaptive Game-Based Math Learning App to Support Personalized Learning and Improve Early Elementary School Students’ Learning. Early. Child. Educ. J.; 2022; pp. 1-16. [DOI: https://dx.doi.org/10.1007/s10643-022-01332-3]

62. Kärki, T.; McMullen, J.; Lehtinen, E. Improving rational number knowledge using the NanoRoboMath digital game. Educ. Stud. Math.; 2022; 1, pp. 101-123. [DOI: https://dx.doi.org/10.1007/s10649-021-10120-6]

63. Faber, T.J.; Dankbaar, M.E.; Kickert, R.; van den Broek, W.W.; van Merriënboer, J.J. Identifying indicators to guide adaptive scaffolding in games. Learn. Instr.; 2022; 83, 101666. [DOI: https://dx.doi.org/10.1016/j.learninstruc.2022.101666]

64. Zeng, L.; Wang, Y. Independent and innovative learning of gymnastics teaching based on computer interaction. Int. J. Elec. Eng. Educ.; 2021; 0020720920984325. [DOI: https://dx.doi.org/10.1177/0020720920984325]

65. Daghestani, L.F.; Ibrahim, L.F.; Al-Towirgi, R.S.; Salman, H.A. Adapting gamified learning systems using educational data mining techniques. Comput. Appl. Eng. Educ.; 2020; 3, pp. 568-589. [DOI: https://dx.doi.org/10.1002/cae.22227]

66. Nilashi, M.; Abumalloh, R.A.; Zibarzani, M.; Samad, S.; Zogaan, W.A.; Ismail, M.Y.; Akib, N.A.M. What Factors Influence Students Satisfaction in Massive Open Online Courses? Findings from User-Generated Content Using Educational Data Mining. Educ. Inf. Technol.; 2022; 7, pp. 1-35. [DOI: https://dx.doi.org/10.1007/s10639-022-10997-7]

67. Lo, J.J.; Shu, P.C. Identification of learning styles online by observing learners’ browsing behaviour through a neural network. Brit. J. Educ. Technol.; 2005; 1, pp. 43-55. [DOI: https://dx.doi.org/10.1111/j.1467-8535.2005.00437.x]

68. Mavroudi, A.; Giannakos, M.; Krogstie, J. Supporting adaptive learning pathways through the use of learning analytics: Developments, challenges and future opportunities. Intract. Learn. Environ.; 2018; 2, pp. 206-220. [DOI: https://dx.doi.org/10.1080/10494820.2017.1292531]

69. Peng, P.; Fu, W. A pattern recognition method of personalized adaptive learning in online education. Mobile. Netw. Appl.; 2022; 3, pp. 1-13. [DOI: https://dx.doi.org/10.1007/s11036-022-01942-6]

70. Roach, M.; Blackmore, P.; Dempster, J.A. Supporting high-level learning through research-based methods: A framework for course development. Innov. Educ. Teach. Int.; 2001; 4, pp. 369-382. [DOI: https://dx.doi.org/10.1080/14703290110074957]

71. Nedungadi, P.; Raman, R. A new approach to personalization: Integrating e-learning and m-learning. ERTD-Educ. Tech. Res.; 2012; 4, pp. 659-678. [DOI: https://dx.doi.org/10.1007/s11423-012-9250-9]

72. Tseng, S.S.; Su, J.M.; Hwang, G.J.; Hwang, G.H.; Tsai, C.C.; Tsai, C.J. An object-oriented course framework for developing adaptive learning systems. Educ. Technol. Soc.; 2008; 2, pp. 171-191.

73. Yaghmaie, M.; Bahreininejad, A. A context-aware adaptive learning system using agents. Expert. Syst. Appl.; 2011; 4, pp. 3280-3286. [DOI: https://dx.doi.org/10.1016/j.eswa.2010.08.113]

74. Wu, L.J.; Chang, K.E. Effect of embedding a cognitive diagnosis into the adaptive dynamic assessment of spatial geometry learning. Intract. Learn. Environ.; 2020; pp. 1-18. [DOI: https://dx.doi.org/10.1080/10494820.2020.1815216]

75. Zhang, J.; Yuan, C.; Wang, C.; Zeng, W.; Dai, S.L. Intelligent adaptive learning and control for discrete-time nonlinear uncertain systems in multiple environments. Neurocomputing; 2021; 462, pp. 31-45. [DOI: https://dx.doi.org/10.1016/j.neucom.2021.07.046]

76. Luo, Z. Using eye-tracking technology to identify learning styles: Behaviour patterns and identification accuracy. Educ. Inf. Technol.; 2021; 4, pp. 4457-4485. [DOI: https://dx.doi.org/10.1007/s10639-021-10468-5]

77. Reinstein, I.; Hill, J.; Cook, D.A.; Lineberry, M.; Pusic, M.V. Multi-level longitudinal learning curve regression models integrated with item difficulty metrics for deliberate practice of visual diagnosis: Groundwork for adaptive learning. Adv. Health. Sci. Educ.; 2021; 3, pp. 881-912. [DOI: https://dx.doi.org/10.1007/s10459-021-10027-0]

78. Wan, H.; Yu, S. A recommendation system based on an adaptive learning cognitive map model and its effects. Intract. Learn. Environ.; 2020; pp. 1-19. [DOI: https://dx.doi.org/10.1080/10494820.2020.1858115]

79. Zourmpakis, A.-I.; Papadakis, S.; Kalogiannakis, M. Education of Preschool and Elementary Teachers on the Use of Adaptive Gamification in Science Education. Int. J. Technol. Enhanc. Learn.; 2022; 14, pp. 1-16. [DOI: https://dx.doi.org/10.1504/IJTEL.2022.120556]

80. Beyhan, S.; Alci, M. Extended fuzzy function model with stable learning methods for online system identification. Int. J. Adapt. Control.; 2011; 2, pp. 168-182. [DOI: https://dx.doi.org/10.1002/acs.1214]

81. Yang, Y.T.C.; Gamble, J.H.; Hung, Y.W.; Lin, T.Y. An online adaptive learning environment for critical-thinking-infused English literacy instruction. Brit. J. Educ. Technol.; 2014; 4, pp. 723-747. [DOI: https://dx.doi.org/10.1111/bjet.12080]

82. Sridharan, S.; Saravanan, D.; Srinivasan, A.K.; Murugan, B. Adaptive learning management expert system with evolving knowledge base and enhanced learnability. Educ. Inf. Technol.; 2021; 5, pp. 5895-5916. [DOI: https://dx.doi.org/10.1007/s10639-021-10560-w]

83. González-Castro, N.; Muñoz-Merino, P.J.; Alario-Hoyos, C.; Kloos, C.D. Adaptive learning module for a conversational agent to support MOOC learners. Australas. J. Educ. Technol.; 2021; 2, pp. 24-44. [DOI: https://dx.doi.org/10.14742/ajet.6646]

84. Agand, P.; Shoorehdeli, M.A.; Khaki-Sedigh, A. Adaptive recurrent neural network with Lyapunov stability learning rules for robot dynamic terms identification. Eng. Appl. Artif. Intel.; 2017; 65, pp. 1-11. [DOI: https://dx.doi.org/10.1016/j.engappai.2017.07.009]

85. Hassan, M.A.; Habiba, U.; Majeed, F.; Shoaib, M. Adaptive gamification in e-learning based on students’ learning styles. Intract. Learn. Environ.; 2021; 4, pp. 545-565. [DOI: https://dx.doi.org/10.1080/10494820.2019.1588745]

86. Binh, H.T.; Trung, N.Q. Responsive student model in an intelligent tutoring system and its evaluation. Educ. Inf. Technol.; 2021; 4, pp. 4969-4991. [DOI: https://dx.doi.org/10.1007/s10639-021-10485-4]

87. Caballé, S.; Mora, N.; Feidakis, M.; Gañán, D.; Conesa, J.; Daradoumis, T.; Prieto, J. CC–LR: Providing interactive, challenging and attractive Collaborative Complex Learning Resources. J. Comput. Assist. Learn.; 2014; 1, pp. 51-67. [DOI: https://dx.doi.org/10.1111/jcal.12021]

88. Chu, H.C.; Chen, J.M.; Kuo, F.R.; Yang, S.M. Development of an adaptive game-based diagnostic and remedial learning system based on the concept-effect model for improving learning achievements in mathematics. Educ. Technol. Soc.; 2021; 4, pp. 36-53.

89. Azevedo, R.; Moos, D.C.; Johnson, A.M.; Chauncey, A.D. Measuring cognitive and metacognitive regulatory processes during hypermedia learning: Issues and challenges. Educ. Psychol.; 2010; 4, pp. 210-223. [DOI: https://dx.doi.org/10.1080/00461520.2010.515934]

90. Lin, Y.T.; Wu, C.C.; Hou, T.Y.; Lin, Y.C.; Yang, F.Y.; Chang, C.H. Tracking students’ cognitive processes during program debugging—An eye-movement approach. IEEE Trans. Educ.; 2015; 3, pp. 175-186. [DOI: https://dx.doi.org/10.1109/TE.2015.2487341]

91. Tempelaar, D. Supporting the less-adaptive student: The role of learning analytics, formative assessment and blended learning. Assess. Eval. High. Educ.; 2020; 4, pp. 579-593. [DOI: https://dx.doi.org/10.1080/02602938.2019.1677855]

92. Narciss, S.; Sosnovsky, S.; Schnaubert, L.; Andrès, E.; Eichelmann, A.; Goguadze, G.; Melis, E. Exploring feedback and student characteristics relevant for personalizing feedback strategies. Comput. Educ.; 2014; 71, pp. 56-76. [DOI: https://dx.doi.org/10.1016/j.compedu.2013.09.011]

93. Sharma, K.; Papamitsiou, Z.; Giannakos, M. Building pipelines for educational data using AI and multimodal analytics: A “grey-box” approach. Br. J. Educ. Technol.; 2019; 6, pp. 3004-3031. [DOI: https://dx.doi.org/10.1111/bjet.12854]

94. Gandomkar, R.; Yazdani, K.; Fata, L.; Mehrdad, R.; Mirzazadeh, A.; Jalili, M.; Sandars, J. Using multiple self-regulated learning measures to understand medical students’ biomedical science learning. Med. Educ.; 2020; 8, pp. 727-737. [DOI: https://dx.doi.org/10.1111/medu.14079]

95. Schumacher, C.; Ifenthaler, D. Investigating prompts for supporting students’ self-regulation–A remaining challenge for learning analytics approaches?. Internet High. Educ.; 2021; 49, 100791. [DOI: https://dx.doi.org/10.1016/j.iheduc.2020.100791]

96. Wang, Y.; Liu, L. Research on the Characteristic Model of Learners in Modern Distance Music Classroom Based on Big Data. Sci. Program; 2022; 2022, 4684461. [DOI: https://dx.doi.org/10.1155/2022/4684461]

97. Adnan, M.; Alarood, A.A.S.; Uddin, M.I.; ur Rehman, I. Utilizing grid search cross-validation with adaptive boosting for augmenting performance of machine learning models. Peerj. Comput. Sci.; 2022; 8, e803. [DOI: https://dx.doi.org/10.7717/peerj-cs.803]

98. Liu, H.I.; Yang, M.N. QoL guaranteed adaptation and personalization in e-learning systems. IEEE Trans. Educ.; 2005; 4, pp. 676-687. [DOI: https://dx.doi.org/10.1109/TE.2005.858398]

99. Wang, F.H. A fuzzy neural network for item sequencing in personalized cognitive scaffolding with adaptive formative assessment. Expert. Syst. Appl.; 2004; 1, pp. 11-25. [DOI: https://dx.doi.org/10.1016/j.eswa.2003.12.001]

100. Karampiperis, P.; Sampson, D. Adaptive learning resources sequencing in educational hypermedia systems. Educ. Technol. Soc.; 2005; 4, pp. 128-147.

101. Santos, J.M.; Anido, L.; Llamas, M.; Álvarez, L.M.; Mikic, F.A. Applying computational science techniques to support adaptive learning. Proceedings of the International Conference on Computational Science (ICCS); St. Petersburg, Russia, 2–4 June 2003; Springer: Berlin/Heidelberg, Germany, 2003; [DOI: https://dx.doi.org/10.1007/3-540-44862-4_116]

102. Tsai, C.J.; Tseng, S.S.; Lin, C.Y. A two-phase fuzzy mining and learning algorithm for adaptive learning environment. Proceedings of the International Conference on Computational Science (ICCS); San Francisco, CA, USA, 28–30 May 2001; Springer: Berlin/Heidelberg, Germany, 2001; [DOI: https://dx.doi.org/10.1007/3-540-45718-6_47]

103. Tseng, J.C.; Chu, H.C.; Hwang, G.J.; Tsai, C.C. Development of an adaptive learning system with two sources of personalization information. Comput. Educ.; 2008; 2, pp. 776-786. [DOI: https://dx.doi.org/10.1016/j.compedu.2007.08.002]

104. Lin, C.F.; Yeh, Y.C.; Hung, Y.H.; Chang, R.I. Data mining for providing a personalized learning path in creativity: An application of decision trees. Comput. Educ.; 2013; 68, pp. 199-210. [DOI: https://dx.doi.org/10.1016/j.compedu.2013.05.009]

105. Ebibi, M.; Fetaji, B.; Fetaji, M. Expert Based Learning (EXBL) Methodology for Developing Mobile Expert Learning Knowledge Management Software System. Tech. Technol. Educ. Manag.; 2012; 2, pp. 864-874.

106. Chang, T.Y.; Ke, Y.R. A personalized e-course composition based on a genetic algorithm with forcing legality in an adaptive learning system. J. Netw. Comput. Appl.; 2013; 1, pp. 533-542. [DOI: https://dx.doi.org/10.1016/j.jnca.2012.04.002]

107. Coello, C.C.; Lechuga, M.S. MOPSO: A proposal for multiple objective particle swarm optimization. Proceedings of the 2002 IEEE Congress on Evolutionary Computation; Honolulu, HI, USA, 12–17 May 2002; Volume 2, pp. 1051-1056. [DOI: https://dx.doi.org/10.1109/CEC.2002.1004388]

108. Martins, A.F.; Machado, M.; Bernardino, H.S.; de Souza, J.F. A comparative analysis of metaheuristics applied to adaptive curriculum sequencing. Soft Comput.; 2021; 16, pp. 11019-11034. [DOI: https://dx.doi.org/10.1007/s00500-021-05836-9]

109. LeCun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature; 2015; 7553, pp. 436-444. [DOI: https://dx.doi.org/10.1038/nature14539]

110. Zhang, L.; Amat, A.Z.; Zhao, H.; Swanson, A.; Weitlauf, A.; Warren, Z.; Sarkar, N. Design of an intelligent agent to measure collaboration and verbal-communication skills of children with autism spectrum disorder in collaborative puzzle games. IEEE Trans. Learn. Technol.; 2020; 3, pp. 338-352. [DOI: https://dx.doi.org/10.1109/TVT.2019.2955959]

111. Dhakshinamoorthy, A.; Dhakshinamoorthy, K. KLSAS—An adaptive dynamic learning environment based on knowledge level and learning style. Comput. Appl. Eng. Educ.; 2019; 2, pp. 319-331. [DOI: https://dx.doi.org/10.1002/cae.22076]

112. Pavlik, P.I.; Eglington, L.G.; Harrell-Williams, L.M. Logistic knowledge tracing: A constrained framework for learner modeling. IEEE Trans. Learn. Technol.; 2021; 5, pp. 624-639. [DOI: https://dx.doi.org/10.1109/TLT.2021.3128569]