Two Decades of Research in e-Learning : A Deep Bibliometric Analysis

E-learning is the electronic version of distance learning: a planned teaching that presupposes a physical separation between the teacher and the student. The distance is either geographic or temporal, and communication can be asynchronous or synchronous, respectively. E-learning has been a target of high-quality research for the last two decades. The purpose of this paper is to analyze the scientific production on e-learning in journals indexed on Elsevier’s Scopus. The sample was composed by 25330 articles from 2000 to 2019. The results obtained by bibliometric analysis showed that rates publication continue to increase. A report was made on the journals, languages, authors, keywords, organizations and countries that publish in the field. This analysis was done for all articles as well as for the most cited articles. The bibliometric analysis was done for a total of 20 years, as well as for four 5-year periods. This article provides information from the past, but mostly clues about research on e-learning in the future.


I. INTRODUCTION
Distance learning is planned teaching that presupposes a physical separation between the teacher and the student. The distance is either geographic -so that the two agents of the teaching-learning process are separated in space -or temporal, and communication can be asynchronous or synchronous, respectively. E-learning is the Internet version as a mean of communication, assuming computer supported collaborative learning (CSCL). E-learning appears at the end of the last century with the growth of the Internet. In 1989, the University of Phoenix became one of the first schools to offer courses on the Internet and introduced FlexNet, which combines online and classroom learning [1]. The concern was whether e-learning was going to replace classrooms [2] or how it would be accepted in in high-tech companies [3]. E-learning was defined as technology-based learning in which learning materials are delivered electronically to remote learners via a computer network [2], a system that includes instruction delivered via all electronic media including the Internet, intranets, extranets, satellite broadcasts, audio/video tape, interactive TV, and CD-ROM [4] or a web-based system that makes information or knowledge available to users or learners and disregards time restrictions or geographic proximity [5]. At the beginning of the century, there were problems with the adoption of distance learning mediated by the Internet, such as technical difficulties, lack of pedagogical models, pre-conceived ideas Manuscript  regarding the old distance learning or even the expectations generated around e-learning. Above all, there were fears: that the teacher's working time would be much longer and even questions related to the ownership of documents and the possibility of copying them. If e-learning is the electronic version of learning, b-learning is blended or mixed [6], m-learning is mobile [7], u-learning is ubiquitous [8], x-learning is the version that includes all the others [9].
Higher education institutions were the privileged places for adopting e-learning. The concern was to measure and test pedagogical and technical methods [10]- [12]. Quickly, e-learning started to be used in secondary schools and the concerns were similar [13]. There was great enthusiasm for the possibilities of distance learning in medical schools [14]. The e-learning platforms, or Learning Management System (LMS), are electronic tools that integrate a large number of interesting applications from the teaching-learning point of view by bringing this grouping quite advantages. Applications have two operating dimensions: asynchronous, not assuming a temporal simultaneity between them (for example, use of electronic mail and discussion forums) and synchronous, in which the activities of the participants take place at the same time and examples of which are online conversations, chats, and systems computer video conference. The discussion was much about technology and the use of learning management systems, like Blackboard [15], MOODLE [16], moving from Blackboard to MOODLE [17] or from MOODLE to Facebook [18].
Some literature reviews were made about some aspects of e-learning: Predictors of E-learning satisfaction in teaching and learning for school teachers [19], the nursing educator's role in e-learning [20] and about the factors influencing e-learning and blended learning in relation to learning outcome, student satisfaction and engagement [21]. Meta-analyzes on the comparison of e-learning with the classroom [22] and the media and pedagogy in undergraduate distance education [23] were also published. Bibliometric analyzes were carried out: 689 journal articles and proceedings retrieved from the Science Citation Index from 2000 to 2008 [24] and 324 articles on workplace e-learning published in academic journals and conference proceedings from 2000 to 2012 [25]. Two decades of quality research in e-learning are celebrated and a bibliometric analysis is necessary.
Bibliometric analysis [26] is the quantitative study of bibliographic material: it provides a general picture of a research field that can be classified by papers, authors and journals. Bibliometric methods employ a quantitative approach for the description, evaluation and monitoring of published research. These methods have the potential to introduce a systematic, transparent and reproducible review process and thus improve the quality of reviews [27]. Bibliometric analysis provides objective criteria that can assess the research development in a field and act as a valuable tool for measuring scholarship quality and productivity [28]. Bibliometric methods offer systematization and replication processes that can improve understanding of the dissemination of knowledge in a field and can highlight gaps and opportunities that contribute to the advancement of the discipline [29].
The aim of this study is to conduct a literature review of the e-learning research using bibliometric methods. The next section presents the research questions. The methodology is defined in the third section. Then the results are presented and at the end they are discussed and a conclusion is made.

II. THE RESEARCH QUESTIONS
The questions, along with the purpose of the review, the intended deliverables and the intended audience, determine how the data are identified, collected and presented [30]. There are several questions that we want to answer in this paper:  How has the evolution of the publication of articles in quality journals related to e-learning been?  What are the characteristics of journals where there is a greater number of publications related to the subject?  What is the approach to e-learning?  Who publishes on the subject? Where do researchers who are interested in e-learning work? What country do they work in? What are the languages in which most of these articles are published?  What are the most cited articles?  What is the purpose of the most cited articles? What is the perspective with which the articles approach the theme?  Who writes the most cited articles? And where do they work? It is intended that all these questions are answered for the total of 20 years, either for four groups of five years periods.

III. METHODOLOGY
The term bibliometrics was first used in 1969 by Alan Pritchard [31], hoping that the term would be explicitly used in all studies which seeked to quantify the processes of written communication and that it would quickly gain acceptance in the field of information science. Moed [32] mentioned the potential of this type of study that reveals the enormous potential of quantitative, bibliometric analyses of the scholarly literature for a deeper understanding of scholarly activity and performance, and highlighted their policy relevance. In scientific research, it is important to get a wider perspective of research already being conducted concerning a relevant subject matter [33] and a bibliometric analysis profile on the research trajectory and dynamics of the research activities across the globe [34]. This is a bibliometric study that systematically analyses the literature using articles indexed at Elsevier's Scopus (Scopus) database. This study conducts a bibliometric analysis of international journal papers that we expect provider a useful reference for future research.   Papers were published in 413 international journals, 11 published of which two hundred or more articles (Table II  Journals    For all publications, the most common WoS subjects are social sciences and computer sciences (0.51 and 0.41), but there are publications in 27 subjects. In the following table (Table IV Subjects, more than 500 articles, 5-year period) we list the subjects with more than 500 articles. There has been a huge increase in the area of medicine and the arts and humanities, in this case the event is especially visible in the last decade.
E-learning, Education, Teaching, Human, Students, Article, Humans, Learning Systems, Distance Learning, Internet and Distance Education are the most common keywords (Fig. 2 Most common keywords.).   (Table V 20 (Table VI Authors with 20 or more articles.). Taiwan is the country affiliation from seven of these 16 authors. Considering country affiliation, the papers are from 168 countries. There are 14 countries that are responsible for at least 500 papers. These 14 countries represent 75% of the total papers ( Fig. 4 Countries with at least 500 papers.).   Open University from United Kingdom is the organization with the most references (306). There are 22 organizations with more than 80 articles (Table VII Organizations with more than 80 references).
Since we had not used the language exclusion criterion, we can now see that English is used in 93% of the articles (Fig. 6. Language articles). There are other 37 languages, like are Spanish 1.6%, German and Chinese 1%. Portuguese and French is used in 0.67% of the articles International Journal of Information and Education Technology, Vol. 11, No. 9, September 2021  In the following table (Table VIII) we list articles with over 500 citations.
In the keywords of the most cited articles we find 4 clusters: C1: Asynchronous learning, computer mediated communications, computer mediated leaning, distance learning, faculty satisfaction, interaction, learning effectiveness, perceived learning, social presence and student satisfaction. C2: data mining, distance education and telelearning, e-learning, e-learning management, evolution of cal system, learner satisfaction and web mining. C3: computer game, engagement, game mechanic, games-based learning, gamification and motivation. C4: elm feature space, elm kernel, ensemble extreme learning machine, incremental learning, online sequential learning, and support vector machine. Kent State University, Purdue University, University of California, University of Plymouth and VA Medical Center are the organization with two of the most cited paper.  Kent State University, Purdue University, University of California, University of Plymouth and VA Medical Center, are the affiliation of two papers each. Using Vosviewer (Fig,  7) we made an analysis of the keywords of the most cited articles on each 5-year period, the authors who published the most each 5-year period and the countries which published the most each period-No significant clusters were find.

V. DISCUSSION AND CONCLUSIONS
The purpose of this paper is to analyze the scientific production on e-learning in journals indexed in Elsevier's Scopus:  How has the evolution of the publication of articles in quality journals related to e-learning been? There is a clear annual growth. This growth is even more  What is the purpose of the most cited articles? What is the perspective with which the articles approach the theme? In the keywords of the most cited articles we find 4 clusters: C1: Asynchronous learning, computer mediated communications, computer mediated leaning, distance learning, faculty satisfaction, interaction, learning effectiveness, perceived learning, social presence and student satisfaction. C2: data mining, distance education and telelearning, e-learning, e-learning management, evolution of cal system, learner satisfaction and web mining. C3: computer game, engagement, game mechanic, games-based learning, gamification and motivation. C4: elm feature space, elm kernel, ensemble extreme learning machine, incremental learning, online sequential learning and support vector machine.  Who writes the most cited articles? And where do they work? J. Duchi, E. Hazan and Y. Singer wrote the most cited paper and B. Babenko, M. H. Yang and S. Belongie wrote the second one. Kent State University, Purdue University, University of California, University of Plymouth and VA Medical Center are the organizations with two of the most cited papers. From the findings, we can see that e-learning continues to be an object of research with a lot of potential and that the best journals still have an interest on the field.