Next Article in Journal
Seasonal Variation in the Element Composition of Dried, Powdered Green Sea Urchin (Strongylocentrotus droebachiensis) from Northern Norway
Previous Article in Journal
Analyzing Managerial Skills for Employability in Graduate Students in Economics, Administration and Accounting Sciences
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 16
Issue 16
10.3390/su16166724
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

Understanding the Growth of Artificial Intelligence in Educational Research through Bibliometric Analysis

by
Ibrahim Delen
1,*,
Nihal Sen
2,
Fatma Ozudogru
3 and
Michele Biasutti
4
1
Mathematics and Science Education Department, University of Usak, 64200 Usak, Türkiye
2
Institute of Educational Sciences, Special Education, Bolu Abant Izzet Baysal University, 14030 Bolu, Türkiye
3
Educational Sciences Department, University of Usak, 64200 Usak, Türkiye
4
Department of Philosophy, Sociology, Education and Applied Psychology, University of Padova, 35139 Padova, Italy
*
Author to whom correspondence should be addressed.
Sustainability 2024, 16(16), 6724; https://doi.org/10.3390/su16166724
Submission received: 15 June 2024 / Revised: 22 July 2024 / Accepted: 2 August 2024 / Published: 6 August 2024
Browse Figures
Versions Notes

Abstract

:
The purpose of this study was to investigate research trends in artificial intelligence studies related to education that were published in the Web of Science database. This review conducted a bibliometric analysis of 4673 articles published between 1975 and 2023 and explored trends in several areas, including the annual distribution of publications, frequently studied topics, top authors, top countries, top universities/departments, top journals and publishers, and top funders. The findings highlighted that the number of studies increased exponentially after 2010. The most often used terms in artificial intelligence research in education were machine learning, deep learning, and data mining. Studies in higher education have been more prevalent than studies in elementary and secondary education. The USA, mainland China, and Australia were the three most productive nations. Most productive authors were connected to academic institutions in Taiwan, Hong Kong, or mainland China. Furthermore, there was little cooperation among the most productive authors andcountries. In addition to the abundance of journals on educational technology, it is crucial to emphasize the dearth of publications on education across different disciplines. To understand how artificial intelligence can support new practices in educational research, interdisciplinary interest and support are needed.

1. Introduction

Science fiction has long entertained the notion of robots mimicking cognitive functions; nevertheless, in more recent times, this notion has evolved from fiction to science, giving rise to the field of artificial intelligence. John McCarthy used the phrase “artificial intelligence” for the first time in 1956 during the famed Dartmouth Conference, where ideas and strategies on artificial intelligence were shared. McCarthy and other academics discussed the possibility of enabling machines to simulate several human intellectual traits, including language comprehension, learning, and self-problem-solving. This idea could be defined as the imitation of human thought processes through utilizing software and computers, and in this way, giving intelligence to machines [1]. In general terms, artificial intelligence is computer systems created for machines to perform human-like cognitive abilities such as learning, reasoning, problem-solving, decision-making, self-correction, and understanding language [2].
Artificial intelligence aims to imitate cognitive abilities through computers and software and has a wide range of applications. One of the most well-known subfields of artificial intelligence is machine learning, which is concerned with deriving algorithms from data [3]. Deep learning uses artificial neural networks to model complicated data structures [4]. Natural language processing investigates how computers comprehend and interact with language [5]. Robotics investigates how machines can move and interact with the real environment [6], while computer vision research examines how machines can analyze and comprehend visual input [7]. The field of knowledge representation and logic deals with the question of how machines can represent knowledge and make logical inferences [8]. Expert systems strive to emulate human expert knowledge in a given domain and apply knowledge in the form of logical rules [9]. Every subfield advances the development of artificial intelligence by delving deeply into new applications of the technology.
These artificial intelligence subfields have the power to fundamentally change a variety of industries and sectors, including manufacturing, healthcare, and entertainment. Artificial intelligence offers unique advantages to students, teachers, and educational institutions with the innovations it brings to the education sector. Artificial intelligence facilitates personalized learning strategies and offers materials and techniques tailored to each student’s needs, learning style, and speed [10]. Learning materials and activities can be tailored to each student’s unique profile by using adaptive learning platforms, which assess students’ strengths and weaknesses. Teachers can also benefit from such research by using them to better understand student performance and give constructive criticism [11,12]. On the other hand, individual learning is not the only way that artificial intelligence may be used in education. Chatbots driven by artificial intelligence provide students with ongoing assistance, helping them with schoolwork, test preparation, and idea comprehension. The process of learning becomes more adaptable and self-governing, providing students with personalized feedback. In addition, artificial intelligence-based instructional technologies help students acquire real-world skills in a variety of fields, such as coding and language learning. Therefore, artificial intelligence could influence how education is scaffolded in the future [13].
The quantity of academic studies on artificial intelligence is a good indicator of the expanding interest in this field. Fifty years ago, authors described artificial intelligence as a “rapidly developing trend” (p. 933) and noted that research in this field had not yet yielded a universal method for formulating complicated problems [14]. The following five years saw a spike in the number of publications, which by 1980 had quickly surpassed 100 per year (the search was conducted using the same keywords presented in the methods section and included the titles, abstracts, and keywords). After surpassing the 1000 yearly publications threshold in 1989, there were more than 10,000 publications in 2000. In 2019, this number had surpassed 100,000 publications annually. There were 1.5 million articles on artificial intelligence in 2023. With about 400,000 publications in the WoS database, Engineering, Electrical, and Electronics is the leading field in artificial intelligence. However, there are a few fields that are receiving less attention. For instance, in the 1985 International Joint Conference on Artificial Intelligence, the topics of artificial intelligence and education received the fewest papers [15]. This trend continued in the following years. Studies published under educational research reached 100 annual publications by 2005 and exceeded 1000 in 2021. Educational research reached nearly 9000 studies by September 2023 in the WoS database. In this study, we will examine artificial intelligence studies published under educational research. In the following section, we will summarize the previous review studies in education.

Artificial Intelligence and Education

Earlier studies predominantly examined artificial intelligence in areas like learning foreign languages. In one of these studies, 41 papers from 2017 to 2023 on using artificial intelligence in language instruction were examined. The authors stated the need for additional empirical research to assess artificial intelligence’s impact on language acquisition [16]. Another study conducting a bibliometric and content analysis of 606 artificial intelligence studies on language acquisition published between 2017 and 2023 revealed a continuous increase in publications through 2022 [17]. A bibliometric examination of 185 studies on artificial intelligence in second language acquisition published between 1995 and 2022 also discussed the potential of AI applications to deliver tailored language learning [18].
Bibliometric analysis studies were also conducted in science education [19]. Researchers explored 76 artificial intelligence studies in science teaching indexed in the Web of Science (WoS) and Scopus databases between 2013 and 2023. There were correlations found between some keywords, including robotics, machine learning, computational thinking, and collaborative learning [19]. In addition, several bibliometric analysis studies investigated the usage of applications [20,21]. Following the COVID-19 pandemic, artificial intelligence chatbots gained growing popularity in education [20]. Another study investigated 93 papers using ChatGPT and underlined the lack of strong relationships between the researchers [21].
Furthermore, an evaluation of the topics, instructional approaches, and learning objectives of 25 studies on artificial intelligence in K–12 education presented how teaching students about artificial intelligence improves their ability to reason ethically, become artificial intelligence-literate, and solve problems [22]. The review stated that learning about AI may promote positive attitudes, enthusiasm for technology, and aspirations for one’s career [22]. While these studies provided a baseline, there is a dearth of research in the literature on artificial intelligence and education [23], and studies that address artificial intelligence in education describe patterns in particular periods and only employ a small number of keywords [24,25]. These constraints created the need for a more in-depth examination of the literature on artificial intelligence in education through bibliometric analyses. In this study, we examined the following research question: What are the patterns in artificial intelligence studies under educational research?

2. Materials and Methods

The growing body of research on artificial intelligence in education indicates that data and conclusions must be gathered, categorized, and interpreted within an overall structure. By highlighting important ideas, authors, areas of concentration for publications, and trends in the literature, bibliometric analyses offer an overview for researchers. Bibliometric analysis employs network analysis and clustering analysis to identify the subjects that are covered in-depth in the literature, the subjects that are left out, the notable writers and academic institutions, and the geographic distribution of the studies. Bibliometric analysis offers a set of tools for evaluating scientific output and identifying research trends [26]. Bibliometric analysis provides a computer-supported analysis of the literature and produces comprehensible visual representations of data using citation information, keywords, and other bibliographic metadata [27]. Bibliometric analysis also determines subgroups and extracts network structures [28]. We used bibliometric analysis to provide an overall picture of the distribution of artificial intelligence in education studies by annual distribution of publications, frequently studied topics, top authors, top countries, leading universities and departments, top publishers, and top funders.
Before conducting this review, we examined the search keywords used in previous studies. In a previous study, 2686 studies published between 2001 and 2021 were found in the WoS database using a search that included only 17 keywords [24]. Using four keywords, another review study searched for studies published from 2011 to 2021 on the Scopus database and found 457 studies [25]. Another study examined 135 studies published in The International Journal of Artificial Intelligence in Education from 2015 to 2019 [29]. The numbers of studies reached in these review studies differ since they examined studies published in specific journals or utilized a limited number of keywords. In this process, the selection of keywords and key phrases is particularly important for effectively analyzing vast amounts of textual content [30]. We used 29 keywords linked to artificial intelligence [31]. These keywords are presented in Figure 1.
After deciding on the keywords, we then proceeded to the search phase. Since the number of journals may vary depending on journal listings over time [28], we conducted our search without limiting it to specific journals. The WoS [24,27,28,29] and Scopus [23,25,26] databases are widely used in bibliometric studies.
As discussed in previous studies, different subject categorizations are offered by the Scopus database (e.g., social science, computer science, decision sciences) [26] and the WoS database (e.g., computer science, information systems, multidisciplinary sciences) [27,28]. In Scopus, computer science is a single category, but the WoS database offers a more detailed classification by dividing computer science into seven categories: computer science artificial intelligence, computer science information systems, computer science theory methods, computer science interdisciplinary applications, computer science software engineering, computer science hardware architecture, and computer science cybernetics. The Scopus database has a category for social sciences, but education is not a sub-category in the Scopus database. The WoS database enabled us to limit our search to specific categories related to education. In addition, searching the WoS database allows researchers to access sophisticated search [27] and categorization [27,28] capabilities. After searching keywords presented in Figure 1 in the WoS database, we included studies under the educational research category, which cover papers on theoretical and practical educational issues. Under educational research, we involved peer-reviewed articles. The search on WoS was performed in September 2023 and 4673 articles were included.
We examined the patterns of artificial intelligence in educational research in seven sections: the annual distribution of publications, frequently studied topics, top authors, top countries, leading universities and departments, top publishers, and top funders. First, we presented the annual distribution of publications by using data from the WoS database. We used VOSviewer (version 1.16.15) to present keyword co-occurrences, author and co-author analyses, and country networks. The quantity of documents in which a keyword appears is indicated by keyword co-occurrences [32]. The total link strength characteristic and the overall strength of a researcher’s co-authorship connections are displayed with bigger nodes [32].
We also used WoS classifications to determine meso and micro classifications [33]. Bibliometric research includes different levels of classification, such as micro, meso, and macro. The micro level is defined as individual academics or members of the scientific community, and the meso level refers to an organization’s activity in the regional aspect that permits the examination of a whole picture of its performance [33]. Citation topics are groups of citations generated algorithmically by the WoS database [34].
To examine the top authors, we presented the authors with the most publications and investigated the author network. The country network illustrates the links among nations [35]. We also presented the top publishing-cited countries. Finally, we reported trends by leading universities, departments, top publishers, journals, and top funders by using data from the WoS database.

3. Results

The trends in the following seven categories are discussed in this section: annual distribution of publications, frequently studied topics, top authors, top countries, leading universities/departments, top journals and publishers, and top funders.

3.1. Annual Distribution of Publications

The WoS database indicates that in 1980, there were over 100 research studies with artificial intelligence-related terms in the title, abstract, or keywords of the study. In contrast, it took almost 30 additional years to reach this threshold in education. The numbers presented in Figure 2 unequivocally demonstrate the field’s expansion. From 1985 to 1999, 126 articles were published. Every year after 2012, there were over 126 articles.

3.2. Frequently Studied Topics

This section displays results from VOSviewer and WoS classifications. We start by demonstrating the co-occurrence of keywords from VOSviewer. VOSviewer groups related objects into distinct clusters, where larger nodes indicate terms that occur more frequently [32]. We will initially display the VOSviewer results, and then the WoS classifications for the included articles.

3.2.1. Keyword Co-Occurrence for Artificial Intelligence in Educational Research

In our study, 10,486 keywords were found. Sixty-five keywords occurred more than 25 times. These keywords were placed in seven different clusters in our analysis. The top ten keywords in Figure 3 were machine learning (occurrence = 430, link strength = 905), deep learning (occurrence = 248, link strength = 475), data mining (occurrence = 245, link strength = 457), educational data mining (occurrence = 243, link strength = 564), learning analytics (occurrence = 221, link strength = 558), higher education (occurrence = 202, link strength = 564), natural language processing (occurrence = 161, link strength = 312), e-learning (occurrence = 132, link strength = 307), artificial intelligence (occurrence = 119, link strength = 290), and data science (occurrence = 98, link strength = 260).
The red cluster has higher education (occurrence = 202, link strength = 564) and deep learning (occurrence = 248, link strength = 475) at the center. The red cluster includes active learning, assessment, case-based reasoning, cognitive science, computational thinking, deep learning, educational technology, engagement, feedback, higher education, learning approaches, mathematics, motivation, problem-solving, problem-based learning, reflection, science education, student engagement, surface learning, and teacher education.
e-learning (occurrence = 132, link strength = 307) is at the center of the green cluster. The green cluster includes collaboration, collaborative learning, distance learning, evaluation, fuzzy logic, genetic algorithm, learning management systems, learning styles, ontology, and recommender systems.
Data mining (occurrence = 245, link strength = 457) and educational data mining (occurrence = 243, link strength = 564) are at the center of the blue cluster. This cluster includes academic performance, artificial neural networks, classification, cluster analysis, clustering, decision trees, distance education, neural networks, and prediction.
Machine learning (occurrence = 430, link strength = 905) is at the center of the orange cluster. This cluster also includes big data, blended learning, COVID-19, data science, MOOCS, natural language processing, online learning, sentiment analysis, and text mining.
The purple cluster has learning strategies (occurrence = 45, link strength = 125) at the center. It also includes data science applications, intelligent tutoring systems, secondary education, self-regulated learning, social media, and teaching/learning strategies.
Artificial intelligence (occurrence = 119, link strength = 290) is at the heart of the black cluster. Automatic speech recognition, mobile learning, pedagogy, and technology are also part of this cluster. The final cluster only includes learning analytics (occurrence = 221, link strength = 558).

3.2.2. WoS Classifications for Artificial Intelligence in Educational Research

The meso-level classification is presented in Table 1. Since we limited our search query to educational research, this classification appeared on top. We presented ten categories following educational research. The following three meso classifications were language and linguistics, knowledge engineering and representation, and management.
The micro-level classification offers a wide range and gives a more detailed analysis of popular topics. The top three groups were self-regulated learning, science education, and learning styles (see Table 2).

3.3. Top Authors

There were 11,467 authors, and only 13 authors had published ten or more articles by September 2023. Table 3 presents the list of these authors. Gwo-Jen Hwang and Wanli Xing are the only authors with more than 20 articles.
In Figure 4, we included 41 authors who had seven or more publications. These authors were grouped into 22 different clusters. There were 13 authors placed individually in a cluster, demonstrating the lack of strong connections among top authors. There were only two clusters presenting a connection among four different authors. Gwo-Jen Hwang is in the blue cluster together with Gary Cheng, Haoran Xie, and Checgjiu Yin. The green cluster includes Gokhan Akcapinar, Hiroaki Ogata, Solomon Sunda Oyelere, and Stephen J. H. Yang. Hiroaki Ogata is among the top ten authors in Table 3. The author with the second-most articles (Wanli Xing) is in the yellow cluster with Bo Pei and Hengtao Tang. The author with the third most articles (Dragan Gasevic) is in the red cluster together with Hendrik Drachsler, Jelena Jovanovic, and Abelardo Pardo.

3.4. Top Countries

There were 114 countries in our analysis. Only eleven countries had more than 100 articles. Table 4 displays the top 10 countries (the Netherlands was placed as the 11th country, with 114 articles). The United States published a little over 25% of the articles and led the citations. Australia ranked third, while mainland China came in second on the list of top countries. Articles published in mainland China received fewer citations than articles from Australia and England.
When we examined the country connections through bibliometric analysis, there were 43 countries with 25 or more articles in Figure 5. The biggest cluster is the red cluster, and it includes England, Germany, Belgium, the Netherlands, Norway, Poland, Portugal, Sweden, Greece, Ireland, Finland, Italy, and Switzerland.
The green cluster includes many South American countries. Spain, Brazil, Chile, Colombia, Ecuador, France, Mexico, and Morocco are in the green cluster. The blue cluster includes Türkiye, India, Russia, Pakistan, South Africa, the United Arab Emirates, and Ukraine. Mainland China, Singapore, Japan, and Taiwan are placed in the black cluster. The United States and Canada are placed in the purple cluster.

3.5. Leading Universities and Departments

There were four universities/university systems in the United States when we investigated the top universities and university systems (see Table 5). Asia and Australia were home to the other universities, with only one located in Europe.
The results presented in Table 5 were transformed when we examined department affiliations (see Table 6). There were no departments from the United States. Departments in the top five are located in mainland China, Hong Kong, and Taiwan.

3.6. Top Journals and Publishers

In this section, we present leading journals and publishers in artificial intelligence education. One out of three articles (33%) in our sample were published in Springer Nature or Taylor and Francis (Table 7).
The top journals have a strong emphasis on educational technology and the top three journals published more than one in seven (15.5%) articles in the sample (Table 8).

3.7. Top Funders

Disparities can be observed in the list of top funders (Table 9). The number of articles financed by the National Science Foundation (231) was greater than the number of articles associated with sponsors in Taiwan, mainland China, and the European Union (194 articles).

4. Discussion

In this review, we used bibliometric analysis to examine peer-reviewed papers on artificial intelligence that were included in the WoS database’s educational research category. VOSviewer software and data from the WoS database were used for data analysis. The first peer-reviewed studies were published in Instructional Science in 1981 [36,37,38]. Although there were some fluctuations in the number of publications between 1990 and 2000, research on artificial intelligence in education saw an upward trend. After 2011, the WoS database published a minimum of 100 publications per year in the category of educational research. In line with these conclusions, additional research revealed that there have been more studies recently [13,39]. Trends and important topics associated with artificial intelligence in education will be covered in this section.

4.1. Trends in Artificial Intelligence in Education

The United States, mainland China, and Australia are the top three countries in this field according to the country-based analysis. It is crucial to note, too, that research published in Australia received more citations than research published in mainland China. Our results show that scientific studies in the field of educational artificial intelligence are increasing in line with similar studies [13,23,40]. Country-based cluster analysis revealed that there were few strong links among authors located on different continents. Countries in Europe, Asia, and North America were grouped in distinct clusters. This conclusion was verified by author analysis, which grouped all 13 authors into separate clusters. Lastly, the United States, Taiwan, mainland China, and the European Union accounted for the top five funding sources for artificial intelligence research in education. The top three notable colleges in our analysis were in the United States [13]. However, the analysis did not include any North American university departments in the top 10. In a similar vein, a different study found that in 2021 and 2022, Chinese authors produced a greater number of articles than American authors [41]. This shifting dynamic may affect how the field develops in the future.
The most prominent authors on artificial intelligence with a focus on educational research were Gwo-Jen Hwang from the National Taiwan University of Science and Technology and Wanli Xing from the University of Florida. In the literature on artificial intelligence, both authors are regarded as prominent figures. Furthermore, the number of publications coming from the United States may be related to the fact that most of the top authors are in the United States. Another study reported a similar result, stating that 15 of the 44 most productive authors are in the United States [13]. These data show that the United States is a leading country in the field of artificial intelligence and that most of the important researchers are based in the United States. On the other hand, the lack of strong connections among top authors was evident from the placement of 13 authors solely in a cluster. Country connections also confirm this result by presenting continent/language based connections.
Journals with a strong interest in educational technology became the leading journals in artificial intelligence research. The data demonstrate that research on artificial intelligence in education is concentrated in a small number of journals, with only six publishing over 100 papers. In addition to the dominance of educational technology journals [40], it is important to underline the lack of disciplinary journals in artificial intelligence under educational research.

4.2. Important Fields of Emphasis in Artificial Intelligence in Education

The top keywords in our analysis were associated with important themes in artificial intelligence research: machine learning [3], deep learning [4], and natural language processing [5]. With its vast potential to better comprehend students’ learning processes, enhance teaching strategies, and raise student accomplishment, the number of machine learning studies continues to rise within the field of education [42]. Large datasets can be analyzed by machine learning algorithms, which can then determine the learning needs and styles of students and provide recommendations for individualized instruction [43]. Evaluation of the efficacy of instructional materials and learning content is another use for machine learning [42]. Deep learning holds a great deal of promise for identifying student behaviors [44]. Research on artificial intelligence aims to create learning environments that suit the needs and preferences of students [45]. Deep learning aims to provide instructors with an automated assessment tool to help with the assessment process. Deep learning algorithms have enhanced the assessment process when compared to traditional methodologies [46,47].
Other prominent keywords (e.g., educational data mining and learning analytics) began to draw the interest of educational scholars since they have enabled teachers to monitor students’ growth comprehensively and make informed decisions [48]. It is noteworthy to mention that most studies used machine learning in higher education [48]. Researchers examined learning analytics and educational data mining techniques for teacher education in a different study [49]. Artificial intelligence in higher education is heavily stressed in both studies [48,49].
In a nutshell, our findings indicated that research in higher education has been conducted more frequently than research in primary and secondary education. Universities have historically covered artificial intelligence in their graduate and undergraduate programs [50]. In addition, the increasing use of artificial intelligence has significant implications for higher education [51]. Furthermore, researchers identified two discourses regarding artificial intelligence. The first is that artificial intelligence is an imperative change, and all stakeholders should respond. The second is artificial intelligence is portrayed as decentering educators and distributing power among employees, students, and machines in higher education [51].
Higher education can benefit from artificial intelligence in several ways, including individualized instruction, timely student response, support for students outside of the classroom, and research assistance [52]. Artificial intelligence has the potential to meet the social and educational needs of students pursuing higher education. Artificial intelligence advances teaching and learning while improving the effectiveness and safety of institutions [53].
It has been proposed that artificial intelligence education could be incorporated into elementary and secondary education due to the growing prevalence of artificial intelligence in many areas of daily life [50]. In addition to preparing students for a society in which artificial intelligence is expected to play a major role, artificial intelligence education can help students become more moral and responsible users of artificial intelligence in the future [54].
Language and linguistics emerged as important themes connected with previous studies. Different aspects of language acquisition are supported by a range of artificial intelligence technologies, including intelligent chatbots, machine translation systems, writing and revision assistants, learning analytics, and adaptive recommendation systems [55,56,57]. Artificial intelligence applications in language instruction have attracted increasing attention [16,17,18], demonstrating their potential to support students’ emotional development, acquisition of knowledge, and language learning progress [39,58].
Using the artificial intelligence software Speeko, students in an experimental study were able to talk more fluently. The computer gave the students customized feedback by assessing their speech patterns, highlighting mistakes, and making suggestions for development [58]. In a different study, the degree of achievement, motivation, and self-regulated learning skills of foreign language learners were all improved by artificial intelligence-mediated foreign language training [59].
Another important theme was e-learning. Using artificial intelligence in e-learning developed into an increasingly important field of study [60]. Most of the e-learning research using artificial intelligence to support assessment and evaluation in e-learning settings focused on the development and application of intelligent tutoring systems.
Finally, science education emerged as a theme in co-occurrence and micro classifications. From 2007 to 2018, authors in the fields of STEM (science, technology, engineering, and mathematics) education and computer science education published numerous articles in higher education [61]. From 2016 to 2022, language learning was the most popular subject in higher education, with computer science education coming in second [41]. Apart from the focus on computer science education, another review study reported 76 science education studies in elementary and secondary grades from 2013 to 2023 [19]. The application of artificial intelligence in elementary and secondary schools has been described by researchers as “shallow integration” (p. 113) [19]. Our findings highlighted the dearth of multidisciplinary connections and placed a strong emphasis on higher education. Only mathematics, science education, and language emerged in our research as important themes, indicating the lack of attention given to other disciplines.

5. Conclusions and Limitations

Artificial intelligence was described as a “rapidly developing trend” (p. 933) in 1975 [14]. Artificial intelligence-related keywords were included in the abstracts of almost 1.5 million publications published in the WoS database over the next fifty years. After 2019, the annual number of publications exceeded 100,000. Engineering, electrical, and electronics is the main field in artificial intelligence, with over 400,000 articles. While one field is leading the way, others are receiving less attention. but some fields are receiving less attention than others. In 1985, The International Joint Conference on Artificial Intelligence included the smallest number of papers linked to artificial intelligence and education [15]. Every year, hundreds of publications are published discussing artificial intelligence in education; however, only 0.05% of artificial intelligence studies are in the field of education in the WoS database.
This study focused on a smaller percentage of artificial intelligence research and originated from the need to provide a more comprehensive analysis for artificial intelligence studies in educational research. There are several limitations to our study, despite it offering a broader perspective with more keywords/key phrases and without time restrictions. Firstly, we incorporated all studies containing keywords associated with artificial intelligence without conducting a content analysis. The increase in the last five years could be linked to the increased number of studies during the pandemic as highlighted by other studies examining the WoS database [62]. It is important to note that the number of articles reached from each institution may vary based on the institutional subscription [62]. Only author-reported keywords and author-reported funding information were used in the analysis. Previous studies acknowledged that funding information gathered from the WoS database may be limited [63] and future studies may delve deep into how funding in different countries may affect the use of artificial intelligence in different countries.
In addition, the number of citation counts provided by the WoS database may vary [64]. Our search was conducted in September 2023, and we did not include all articles published in 2023. We analyzed articles from only one database. Nonetheless, we hope that the disparities (e.g., the paucity of research across various disciplines, the lack of interdisciplinary connections, the lack of connections between leading authors and countries, the differences in departmental and university levels, and the changing dynamics, i.e., Asian departments publishing more papers in recent years) presented in this review will provide educational policymakers and scholars with useful ideas to optimize the benefits from the body of knowledge already available in this field. We hope our results will encourage an increase in the number of artificial intelligence applications in elementary and secondary schools in different disciplines.

Author Contributions

I.D. took charge of all sections of the manuscript. N.S. worked on the introduction and discussion. F.O. contributed to the Methods, Results, and Discussion. M.B. worked on the Discussion and editing the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

The data used in this study were downloaded from the Web of Science database.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Konar, A. Artificial Intelligence and Soft Computing: Behavioral and Cognitive Modeling of the Human Brain; CRC Press: Boca Raton, FL, USA, 2018. [Google Scholar]
  2. Garg, P.K. Overview of artificial intelligence. In Artificial Intelligence: Technologies, Applications, and Challenges; Sharma, L., Garg, P.K., Eds.; Chapman & Hall: London, UK, 2021; pp. 2–17. [Google Scholar]
  3. Korkmaz, C.; Correia, A.P. A review of research on machine learning in educational technology. Educ. Media Int. 2019, 56, 250–267. [Google Scholar] [CrossRef]
  4. Shanthamallu, U.S.; Spanias, A. Machine and Deep Learning Algorithms and Applications; Morgan & Claypool Publishers: San Rafael, CA, USA, 2021. [Google Scholar]
  5. Lopez-Martinez, R.E.; Sierra, G. Research trends in the international literature on natural language processing, 2000–2019—A bibliometric study. J. Scientometr. Res. 2000, 9, 310–318. [Google Scholar] [CrossRef]
  6. Raj, M.; Seamans, R. Primer on artificial intelligence and robotics. J. Organ. Des. 2019, 8, 11. [Google Scholar] [CrossRef]
  7. Jin, L.; Tan, F.; Jiang, S. Generative adversarial network technologies and applications in computer vision. Comput. Intell. Neurosci. 2020, 2020, 1459107. [Google Scholar] [CrossRef] [PubMed]
  8. Bench-Capon, T.J. Knowledge Representation: An Approach to Artificial Intelligence; Elsevier: Amsterdam, The Netherlands, 2014. [Google Scholar]
  9. Gupta, I.; Nagpal, G. Artificial Intelligence and Expert Systems; Mercury Learning and Information: Herndon, VA, USA, 2020. [Google Scholar]
  10. Chatterjee, S.; Bhattacharjee, K.K. Adoption of artificial intelligence in higher education: A quantitative analysis using structural equation modeling. Educ. Inf. Technol. 2020, 25, 3443–3463. [Google Scholar] [CrossRef]
  11. Fahimirad, M.; Kotamjani, S.S. A review on application of artificial intelligence in teaching and learning in educational contexts. Int. J. Learn. Dev. 2018, 8, 106–118. [Google Scholar] [CrossRef]
  12. Gligorea, I.; Cioca, M.; Oancea, R.; Gorski, A.T.; Gorski, H.; Tudorache, P. Adaptive learning using artificial intelligence in e-learning: A literature review. Educ. Sci. 2023, 13, 1216. [Google Scholar] [CrossRef]
  13. Chen, X.; Xie, H.; Hwang, G.J. A multi-perspective study on artificial intelligence in education: Grants, conferences, journals, software tools, institutions, and researchers. Comput. Educ. Artif. Intell. 2020, 1, 100005. [Google Scholar] [CrossRef]
  14. Tikhomirov, O.K. Philosophical and psychological problems of artificial intelligence. IJCAI 1975, 932–937. [Google Scholar]
  15. Good, R. Artificial intelligence and science education. J. Res. Sci. Teach. 1987, 24, 325–342. [Google Scholar] [CrossRef]
  16. Law, L. Application of generative artificial intelligence (GenAI) in language teaching and learning: A scoping literature review. Comput. Educ. Open 2024, 6, 100174. [Google Scholar] [CrossRef]
  17. Rahman, A.; Raj, A.; Tomy, P.; Hameed, M.S. A comprehensive bibliometric and content analysis of artificial intelligence in language learning: Tracing between the years 2017 and 2023. Artif. Intell. Rev. 2024, 57, 107. [Google Scholar] [CrossRef]
  18. Kartal, G.; Yesilyurt, Y.E. A bibliometric analysis of artificial intelligence in L2 teaching and applied linguistics between 1995 and 2022. ReCALL 2024, 1–17. [Google Scholar] [CrossRef]
  19. Jia, F.; Sun, D.; Looi, C.K. Artificial intelligence in science education (2013–2023): Research trends in ten years. J. Sci. Educ. Technol. 2024, 33, 94–117. [Google Scholar] [CrossRef]
  20. Lin, Y.; Yu, Z. A bibliometric analysis of artificial intelligence chatbots in educational contexts. Interact. Technol. Smart Educ. 2024, 21, 189–213. [Google Scholar] [CrossRef]
  21. Pradana, M.; Elisa, H.P.; Syarifuddin, S. Discussing ChatGPT in education: A literature and bibliometric analysis. Cogent Educ. 2023, 10, 2243134. [Google Scholar] [CrossRef]
  22. Lee, S.J.; Kwon, K. A systematic review of AI education in K-12 classrooms from 2018 to 2023: Topics, strategies, and learning outcomes. Comput. Educ. Artif. Intell. 2024, 6, 100211. [Google Scholar] [CrossRef]
  23. Song, P.; Wang, X. A bibliometric analysis of worldwide educational artificial intelligence research development in recent twenty years. Asia Pac. Educ. Rev. 2020, 21, 473–486. [Google Scholar] [CrossRef]
  24. Talan, T. Artificial intelligence in education: A bibliometric study. Int. J. Res. Educ. Sci. 2021, 7, 822–837. [Google Scholar] [CrossRef]
  25. Prahani, B.K.; Rizki, I.A.; Jatmiko, B.; Suprapto, N.; Amelia, T. Artificial intelligence in education research during the last ten years: A review and bibliometric study. Int. J. Emerg. Technol. Learn. 2022, 17, 169–188. [Google Scholar] [CrossRef]
  26. Cascajares, M.; Alcayde, A.; Salmerón-Manzano, E.; Manzano-Agugliaro, F. The bibliometric literature on Scopus and WoS: The medicine and environmental sciences categories as case of study. Int. J. Environ. Res. Public Health 2021, 18, 5851. [Google Scholar] [CrossRef]
  27. Tomaszewski, R. Visibility, impact, and applications of bibliometric software tools through citation analysis. Scientometrics 2023, 128, 4007–4028. [Google Scholar] [CrossRef]
  28. McGuigan, G.S.; Morçöl, G.; Grosser, T.A. Social network analysis of academic journals in public administration in the early twenty-first century: Examining journal level bibliometrics with network analysis. Scientometrics 2023, 128, 6561–6588. [Google Scholar] [CrossRef]
  29. Baek, C.; Doleck, T. A bibliometric analysis of the papers published in the Journal of Artificial Intelligence in Education from 2015–2019. Int. J. Learn. Anal. Artif. Intell. Educ. 2020, 2, 67–84. [Google Scholar] [CrossRef]
  30. Siddiqi, S.; Sharan, A. Keyword and keyphrase extraction techniques: A literature review. Int. J. Comput. Appl. 2015, 109, 18–23. [Google Scholar] [CrossRef]
  31. Bawack, R.E.; Wamba, S.F.; Carillo, K.D.A.; Akter, S. Artificial intelligence in e-commerce: A bibliometric study and literature review. Electron. Mark. 2022, 32, 297–338. [Google Scholar] [CrossRef]
  32. Van Eck, N.J.; Waltman, L. VOSviewer Manual; Univeristeit Leiden: Leiden, The Netherlands, 2019. [Google Scholar]
  33. Ibrahim, S.K. Scientometrics Assessment of Malaysian Social Science Research in the Web of Science (2007–2017). Doctoral Dissertation, University of Malaya, Kuala Lumpur, Malaysia, 2021. [Google Scholar]
  34. Clarivate. Citation Topics. Available online: https://incites.help.clarivate.com/Content/Research-Areas/citation-topics.htm (accessed on 12 December 2023).
  35. Oladinrin, O.T.; Arif, M.; Rana, M.Q.; Gyoh, L. Interrelations between construction ethics and innovation: A bibliometric analysis using VOSviewer. Constr. Innov. 2023, 23, 505–523. [Google Scholar] [CrossRef]
  36. Gallagher, J.P. The effectiveness of man-machine tutorial dialogues for teaching attribute blocks problem-solving skills with an artificial intelligence CAI system. Instr. Sci. 1981, 10, 297–332. [Google Scholar] [CrossRef]
  37. Priest, A.G. Artificial intelligence and learning: Conference reports. Instr. Sci. 1981, 10, 277–285. [Google Scholar] [CrossRef]
  38. Willems, J. Problem-based (group) teaching: A cognitive science approach to using available knowledge. Instr. Sci. 1981, 10, 5–21. [Google Scholar] [CrossRef]
  39. Wang, X.; Liu, Q.; Pang, H.; Tan, S.C.; Lei, J.; Wallace, M.P.; Li, L. What matters in AI-supported learning: A study of human-AI interactions in language learning using cluster analysis and epistemic network analysis. Comput. Educ. 2023, 194, 104703. [Google Scholar] [CrossRef]
  40. Moreno-Guerrero, A.J.; Lopez-Belmonte, J.; Marin-Marin, J.A.; Soler-Costa, R. Scientific development of educational artificial intelligence in Web of Science. Future Internet 2020, 12, 124. [Google Scholar] [CrossRef]
  41. Crompton, H.; Burke, D. Artificial intelligence in higher education: The state of the field. Int. J. Educ. Technol. High. Educ. 2023, 20, 22. [Google Scholar] [CrossRef]
  42. Tiwari, R. The integration of AI and machine learning in education and its potential to personalize and improve student learning experiences. Int. J. Sci. Res. Eng. Manag. 2023, 7, 1–11. [Google Scholar] [CrossRef]
  43. Wu, J.Y.; Hsiao, Y.C.; Nian, M.W. Using supervised machine learning on large-scale online forums to classify course-related Facebook messages in predicting learning achievement within the personal learning environment. Interact. Learn. Environ. 2018, 28, 65–80. [Google Scholar] [CrossRef]
  44. Aung, A.M.; Ramakrishnan, A.; Whitehill, J. Who are they looking at? Automatic eye gaze following for classroom observation video analysis. In Proceedings of the 11th International Conference on Educational Data Mining, Buffalo, NY, USA, 15–18 July 2018. [Google Scholar]
  45. Abhinav, K.; Subramanian, V.; Dubey, A.; Bhat, P.; Venkat, A.D. Lecore: A framework for modeling learner’s preference. In Proceedings of the 11th International Conference on Educational Data Mining, Buffalo, NY, USA, 15–18 July 2018. [Google Scholar]
  46. Alvarado, J.G.; Ghavidel, H.A.; Zouaq, A.; Jovanovic, J.; McDonald, J. A comparison of features for the automatic labeling of student answers to open-ended questions. In Proceedings of the 11th International Conference on Educational Data Mining, Buffalo, NY, USA, 15–18 July 2018. [Google Scholar]
  47. Sales, A.; Botelho, A.; Patikorn, T.; Heffernan, N.T. Using big data to sharpen design-based inference in A/B tests. In Proceedings of the 11th International Conference on Educational Data Mining, Buffalo, NY, USA, 15–18 July 2018. [Google Scholar]
  48. Du, X.; Yang, J.; Hung, J.L.; Shelton, B. Educational data mining: A systematic review of research and emerging trends. Inf. Discov. Deliv. 2020, 48, 225–236. [Google Scholar] [CrossRef]
  49. Salas-Pilco, S.Z.; Xiao, K.; Hu, X. Artificial intelligence and learning analytics in teacher education: A systematic review. Educ. Sci. 2022, 12, 569. [Google Scholar] [CrossRef]
  50. Rizvi, S.; Waite, J.; Sentance, S. Artificial Intelligence teaching and learning in K-12 from 2019 to 2022: A systematic literature review. Comput. Educ. Artif. Intell. 2023, 4, 100145. [Google Scholar] [CrossRef]
  51. Bearman, M.; Ryan, J.; Ajjawi, R. Discourses of artificial intelligence in higher education: A critical literature review. High. Educ. 2023, 86, 369–385. [Google Scholar] [CrossRef]
  52. Dhawan, S.; Batra, G. Artificial intelligence in higher education: Promises, perils, and perspective. OJAS Expand. Knowl. Horiz. 2020, 11, 11–22. [Google Scholar]
  53. Kuleto, V.; Ilic, M.; Dumangiu, M.; Rankovic, M.; Martins, O.M.D.; Paun, D.; Mihoreanu, L. Exploring opportunities and challenges of artificial intelligence and machine learning in higher education institutions. Sustainability 2021, 13, 10424. [Google Scholar] [CrossRef]
  54. Miao, F.; Shiohira, K. K-12 AI Curricula: A Mapping of Government-Endorsed AI Curricula; The United Nations Educational, Scientific and Cultural Organization: Paris, France, 2022. [Google Scholar]
  55. Pokrivcakova, S. Preparing teachers for the application of AI-powered technologies in foreign language education. J. Lang. Cult. Educ. 2019, 7, 135–153. [Google Scholar] [CrossRef]
  56. Vo, A.; Nguyen, H. Generative artificial intelligence and ChatGPT in language learning: EFL students’ perceptions of technology acceptance. J. Univ. Teach. Learn. Pract. 2024, 21, 1–19. [Google Scholar] [CrossRef]
  57. Wang, T.; Lund, B.D.; Marengo, A.; Pagano, A.; Mannuru, N.R.; Teel, Z.A.; Pange, J. Exploring the potential impact of artificial intelligence (AI) on international students in higher education: Generative AI, chatbots, analytics, and international student success. Appl. Sci. 2023, 13, 6716. [Google Scholar] [CrossRef]
  58. Shafiee Rad, H. Revolutionizing L2 speaking proficiency, willingness to communicate, and perceptions through artificial intelligence: A case of Speeko application. Innov. Lang. Learn. Teach. 2024, 1–16. [Google Scholar] [CrossRef]
  59. Wei, L. Artificial intelligence in language instruction: Impact on English learning achievement, L2 motivation, and self-regulated learning. Front. Psychol. 2023, 14, 1261955. [Google Scholar] [CrossRef]
  60. Tang, K.Y.; Chang, C.Y.; Hwang, G.J. Trends in artificial intelligence-supported e-learning: A systematic review and co-citation network analysis (1998–2019). Interact. Learn. Environ. 2023, 31, 2134–2152. [Google Scholar] [CrossRef]
  61. Zawacki-Richter, O.; Marín, V.I.; Bond, M.; Gouverneur, F. Systematic review of research on artificial intelligence applications in higher education—Where are the educators? Int. J. Educ. Technol. High. Educ. 2019, 16, 39. [Google Scholar] [CrossRef]
  62. Liu, F. Retrieval strategy and possible explanations for the abnormal growth of research publications: Re-evaluating a bibliometric analysis of climate change. Scientometrics 2023, 128, 853–859. [Google Scholar] [CrossRef]
  63. Liu, W.; Tang, L.; Hu, G. Funding information in Web of Science: An updated overview. Scientometrics 2020, 122, 1509–1524. [Google Scholar] [CrossRef]
  64. Hu, G.; Wang, L.; Ni, R.; Liu, W. Which h-index? An exploration within the Web of Science. Scientometrics 2020, 123, 1225–1233. [Google Scholar] [CrossRef]
Sustainability 16 06724 g001
Figure 1. Keywords and inclusion of studies.
Figure 1. Keywords and inclusion of studies.
Sustainability 16 06724 g001
Sustainability 16 06724 g002
Figure 2. Distribution of studies by year.
Figure 2. Distribution of studies by year.
Sustainability 16 06724 g002
Sustainability 16 06724 g003
Figure 3. Keyword co-occurrence.
Figure 3. Keyword co-occurrence.
Sustainability 16 06724 g003
Sustainability 16 06724 g004
Figure 4. Co-author analysis.
Figure 4. Co-author analysis.
Sustainability 16 06724 g004
Sustainability 16 06724 g005
Figure 5. Country network.
Figure 5. Country network.
Sustainability 16 06724 g005
Table 1. WoS meso classifications.
Table 1. WoS meso classifications.
#Meso ClassificationsNumber of Publications#Meso ClassificationsNumber of Publications
1Education and Educational Research22936Artificial Intelligence and Machine Learning146
2Language and Linguistics3207Social Psychology116
3Knowledge Engineering Representation3058Communication85
4Management1829Nursing63
5Neuroscanning15010Computer Vision and Graphics45
Table 2. WoS micro classifications.
Table 2. WoS micro classifications.
#Micro ClassificationsNumber of Publications#Micro ClassificationsNumber of Publications
1Self-Regulated Learning11606Natural Language Processing107
2Science Education2457Collaborative Filtering105
3Learning Styles2438Computational Thinking98
4Language Policy1839Phonological Awareness91
5Teacher Education11510Technology Acceptance Model80
Table 3. Top authors.
Table 3. Top authors.
#AuthorsNumber of Publications#AuthorsNumber of Publications
1Hwang, Gwo-Jen225Baker, Ryan12
2Xing, Wanli206Rose, Carolyn11
3Gasevic, Dragan18 Chen, Chih-Ming
4Tawfik, Andrew A.
Crossley, Scott
Salas-Rueda, Ricardo-Adán		167Zhai, Xiaoming
Drachsler, Hendrik
Ogata, Hiroaki
Shuang, Li		10
Table 4. Top countries/regions.
Table 4. Top countries/regions.
#CountryNumber of PublicationsNumber of Citations
1USA119821,614
2Mainland China5263792
3Australia3285677
4England2885581
5Spain2523671
6Taiwan2334123
7Canada2143376
8Türkiye1661656
9Germany1452602
10India134889
Table 5. Leading universities and university systems.
Table 5. Leading universities and university systems.
#AffiliationsNumber of Publications#AffiliationsNumber of Publications
1State University System of Florida995University of London55
2University System of Georgia845University of Hong Kong55
3University of California System608Monash University52
4Pennsylvania Commonwealth System of Higher Education599National Taiwan Normal University47
5Beijing Normal University5510Nanyang University46
Table 6. Leading departments.
Table 6. Leading departments.
#Affiliations with DepartmentNumber of Publications
1Beijing Normal University Faculty of Education62
2The University of Hong Kong Faculty of Education59
3The Chinese University of Hong Kong Faculty of Education39
4Beijing Normal University School of Educational Technology38
5National Taiwan University of Science and Technology Graduate Institute of Digital Learning and Education36
6The Chinese University of Hong Kong Department of Curriculum and Instruction35
7National Central University College of Electrical Engineering and Computer Science32
8The Education University of Hong Kong Faculty of Liberal Arts and Social Sciences27
9The Education University of Hong Kong Department of Mathematics and Information Technology
The University of Edinburg College of Science and Engineering		26
10Kyoto University Academic Centre for Computing and Media Studies25
Table 7. Top publishers.
Table 7. Top publishers.
#PublisherNumber of Publications#PublisherNumber of Publications
1Springer Nature8006Kassel Univ Press198
2Taylor and Francis7437IEEE146
3Elsevier3648Inderscience Enterprises Ltd129
4Wiley3099IGI Global106
5Sage20510Emerald Group Publishing104
Table 8. Top journals.
Table 8. Top journals.
#JournalNumber of Publications
1Education and Information Technologies266
2International Journal of Emerging Technologies
in Learning		249
3Computers Education210
4IEEE Transactions on Learning Technologies129
5Interactive Learning Environments111
6Educational Technology Society109
7British Journal of Educational Technology78
8Education Sciences76
9Journal of Computer Assisted Learning60
10International Journal of Continuing Engineering59
Education and Lifelong Learning
Table 9. Top funders.
Table 9. Top funders.
#FunderNumber of Publications#FunderNumber of Publications
1National Science Foundation (NSF)1756US Department of Education37
2National Natural Science Foundation of China797Spanish Government36
3Ministry of Science and Technology Taiwan668Ministry of Education Culture Sports Science and Technology Japan33
4NSF Directorate for STEM Education569Japan Society for the Promotion of Science29
5European Union4910Australian Research Council24
Grants in Aid for Scientific Research Kakenhi (Japan)24
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.

Share and Cite

MDPI and ACS Style

Delen, I.; Sen, N.; Ozudogru, F.; Biasutti, M. Understanding the Growth of Artificial Intelligence in Educational Research through Bibliometric Analysis. Sustainability 2024, 16, 6724. https://doi.org/10.3390/su16166724

AMA Style

Delen I, Sen N, Ozudogru F, Biasutti M. Understanding the Growth of Artificial Intelligence in Educational Research through Bibliometric Analysis. Sustainability. 2024; 16(16):6724. https://doi.org/10.3390/su16166724

Chicago/Turabian Style

Delen, Ibrahim, Nihal Sen, Fatma Ozudogru, and Michele Biasutti. 2024. "Understanding the Growth of Artificial Intelligence in Educational Research through Bibliometric Analysis" Sustainability 16, no. 16: 6724. https://doi.org/10.3390/su16166724

APA Style

Delen, I., Sen, N., Ozudogru, F., & Biasutti, M. (2024). Understanding the Growth of Artificial Intelligence in Educational Research through Bibliometric Analysis. Sustainability, 16(16), 6724. https://doi.org/10.3390/su16166724

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop