Artificial intelligence in education

AI performance prediction modeling is widely used to identify at-risk students and consequently establish student-centered learning pathways, and optimize instructional design and development (Ouyang et al., 2023). Alternatively, it is necessary to reflect on AI, chatbots, and plagiarism in higher education (King & chatGPT, 2023). The above tools, if misapplied, will not develop knowledge, professional competencies, or critical thinking skills, which are vital in diverse professions.

The personalized intelligent manufacturing paradigm with cognitive automation capabilities (Li et al., 2023) has attracted increasing attention by offering an adaptable and flexible solution. The paradigm leverages the advantages of both deep neural networks and reinforcement learning by adopting the power of representation learning, to make accurate and fast decisions when faced with dynamic and complex situations.

This intelligence system develops foresight science with a systematic approach to generate future predictions in planning and management using analytical and predictive tools to understand the past and present (Ednie et al., 2023), while providing information on the future of education. The adoption of appropriate practices using AI-enabled techniques facilitates teaching and learning performance in higher education (Hua-Hu, 2023).

The analysis of learning tactics and strategies in an online environment (Lahza et al., 2023) is a form of pedagogical support for collaborating with students in the development of academic-related content. Analysis of learning tactics is emerging as a viable approach to engage students in higher-order learning. In all the above instances, ethical considerations of AI must be considered (Sun & Ye, 2023). One of the ultimate challenges in philosophy is therefore, determining what deserves moral regard.

Digital explosion and entrepreneurship education is having an impact on promoting entrepreneurial intent for business students (Dabbous & Boustani, 2023) and thus, empowering AI as a form of entrepreneurship. Adoption of AI-based teaching robots (T-bots) for learning using the technology adoption model and context-specific variables (Hussain, 2023) aligns with education efforts in fostering intelligent decisions.

It has become essential for today's learners to gain basic literacy and AI competencies, yet educators do not care about these era-defining trends in contemporary pedagogy (Chai et al., 2023). According to Zhen et al. (2023), virtual technology is widely employed in various sectors such as personal assistance, intelligent customer service, online in various educational modalities and for assessing physical activity in high school children (Ahmed et al., 2023), data visualization, augmented reality, and virtual reality (Cerqueira et al., 2023).

An informed perspective of learning through AI raises concerns on how to train users in handling changes brought about by the introduction of the technology (Arbelaez-Ossa et al., 2023; Ursani & Ursani, 2023). The future of standardized assessment, validity, and confidence in assessment and scoring algorithms (Aloisi, 2023), therefore considers challenges of using AI to aid standardized high-risk assessment.

Artificial intelligence software applied in education

Among the AI tools applied in online education are special types of software. English teaching and learning in this information age should be carried out with the help of intelligent system software to counter the disadvantages of the traditional system of teaching and learning (Dong, 2022). According to the current circumstances of teaching in addition to learning, semantic Web technology and AI technology are combined to build an advanced system of teaching and learning.

In the context of AI, scientific research practice has been simplified in terms of technological form and content, that limits students' ability to comprehensively apply technology. The above-mentioned scenario, therefore, becomes an obstacle in the development of superior technological features in the context of AI (Hu et al., 2023). Online and adaptive learning has already advanced within academia. However, the most significant changes are yet to come. These evolving technologies have the potential to change traditional roles in colleges and universities to the point where many educators will reconsider their roles as teachers, researchers, and administrators (Picciano, 2019).

The future of higher education and education through online technology, specifically adaptive learning and analytics infused by AI software, is increasingly developing as part of teaching methodology. Transformation of educational concepts, reform of teaching courses, and reorganization of teaching materials are critical ways to strengthen the quality of contemporary education through AI (Zhou, 2023). This is related to the development of teaching materials suitable for various regions and schools to deepen students' understanding and enhance their interest in courses.

It also serves to make decisions in application domains such as justice, health care, and education (Hort et al., 2023) and as deep learning systems trained to read, classify, and respond to human emotions (Ho et al., 2023). Computer-assisted instruction also has some degree of lock-in. The use of AI causes students to learn passively, without goals, education, and guidance (Shen & Tan, 2023). Based on the above scenarios, there are models for evaluating the effectiveness of AI-based computer-assisted physical education teaching and training, which help to focus goals and determine guidance.

Real object understanding using 3D haptic virtual reality for online education is another tool (Allaoua-Chelloug et al., 2023). This computer vision technology incorporates tasks that implement basic filtering for image classification. The main areas of research in this field include object detection and recognition, and interesting aspects of education at different levels.

Detailed exploration of AI and digital education provides a sustainable impact on youth and society (Tongkachok et al., 2023). Therefore, using AI launches some digital applications for digital education, an element that increased during the COVID-19 pandemic. For all the above instances, it is necessary to define the research question: What are the incidences of AI in contemporary education through a bibliometric analysis? Establishing as a general objective: To analyze bibliometrically the incidences of AI in contemporary education.

Instrument for data collection

For the research, conceptual categories are established considering the incidence of AI in education, from which the dimensions (D) and indicators (I) that lead the research are defined and are based on the analysis of the previously established information:

• D.1: Authors. I: Total citations, citations per year, publication timeliness, affiliation, country, H index.

• D.2: Journals. I: Main publications, journal quartiles, indexation, impact factor, and countries.

• D.3: Contributions I: Subject matter, methodologies used, analysis techniques, and samples used.

Population and sample

The steps carried out for the visual design of the PRISMA diagram (Figure 1) are described below.

1) Identification of the key steps of the systematic review: the systematic review generally includes sub-steps such as:

• Search for studies: The following search equation was used: title ("artificial and intelligence and + and education") and limit-to (language, "English") or limit-to (language, "Spanish"). As can be seen in the equation, English and Spanish were the inclusion criteria.

• Selection of studies: To obtain the best results in the exploration carried out, search strategies were applied to all the literature indexed in Scopus, Mendeley, and Science Direct related to this topic up to 2023.

• Data extraction: The tools of the bibliometric package of the R statistical program (Aria & Cuccurullo, 2017) were used to extract the information of the variables to be analyzed according to the dimensions explained above.

• Evaluation of the quality of the evidence: After the previous steps, the results were analyzed descriptively, and based on this information, the tables and graphs presented in the results section were prepared.

D.1: Authors

One of the indicators of AI in education is the authors who publish most on the subject, as they depict a direct relationship with contemporary topics on the subject. The authors' dimension highlights the people who research and write about the subject under research. Their objective is to share their findings and discoveries with the scientific community and the public and to contribute to the advancement of knowledge in their field of study. With this statement in mind, we will analyze some indicators of AI in contemporary education that directly impact the authors (Table 1). The table shows the main authors publishing in journals focused on AI topics globally. Only 10 authors were selected using Bibliometrix. Also present are the total number of citations per year, the most current AI-related articles in education, the H index that provides the citation level, affiliation, and countries. In order to achieve the greatest possible effectiveness, these variables were reviewed individually by each author in their own scientific network platforms such as Google Scholar, Scopus index, among others,

However, the analysis here shows that developed countries are still at the forefront of technologies and therefore of AI in education. Likewise, the authors are in the best universities in their countries that provide them with a high possibility of research and time to carry it out. It is an extraordinarily topical subject considering the years in question. Citation rates are high, which shows the interest in the subject, and this has allowed the growth of the researchers mentioned in Table 1.

North America, Europe, and Oceania lead, followed by the Asian continent. The total absence of South America and Africa continues to demonstrate that the generation of AI for education is a matter of investment, resources, and budgetary availability. It is noteworthy that these authors established their publications until the year 2020 and although they are still among the first, there is no evidence of publications between 2021 and 2023. In general, there are 10 authors; eight of them belong to developed countries and two to underdeveloped countries. The number of publications within the last five years is seven and three within the last 10 years. Likewise, the mean overall average citation index is 113, the mean citation index per year is 20 and the mean H-index is 19. The leading country in AI in contemporary education according to its authors is Australia (2), considering the top 10 authors.

D.2: Journals

A scientific journal is a periodical that contains original, peer-reviewed articles in a specific field of science. Scientific journals are a means of disseminating recent research and discoveries to a community of researchers and specialists in a subject and are an important part of the review and validation process in science. For all the above reasons, it is necessary to establish the essential elements of the main journals that publish topics related to AI in contemporary education as a dimension of research (Table 2).

Considering Table 2, it can be stated that within the 10 scientific journals that publish the most on the topic of AI in education, 80% are in the first quartile of publication (Q-1) and 20% in the second quartile (Q-2). Ninety percent belong to the Journal Citation Reports (JCR) and 10% to the Scimago Journal & Country Rank (SJR). The above defines a position that all publications related to the research are high impact and evaluated by double-blind peer review systems, Open Access, and are collected in world-class journals compiled by the Scopus database. The average impact factor of the journals is 64.2.

Countries publishing related reputable journals include the United States of America that is in first place with six journals, while Singapore, China, Switzerland, and the United Kingdom all have one journal. All the publications that promote the topic of AI in contemporary education are from developed countries centered on three continents: North America, Europe, and Asia. South America and Africa are absent in journals of this type. This last indicator coincides with the same indicator of the authors' countries, which solidifies that AI in education is a topic that is promoted by developed countries.

D.3: Contributions

Scientific contributions are original and significant contributions that scientists make in their areas of research to expand knowledge and solve problems. These contributions may include discoveries, theories, methods, innovative techniques, experiments, and studies that confirm or refute hypotheses. The scientific contributions of AI in contemporary education are critical for the progress of science and the improvement of the quality of life, so we analyze in this research (Table 3) the main impacts of these methodologies and techniques.

All contributions revolve around AI in online education. However, similarities and differences among them are enhanced. For example, employing AI in teaching and learning, in the evolution and revolution of education, and the insertion of robots in different educational areas through dynamic modeling. Contemporary trends are developing such as educational robots (T-bots) and intelligent classrooms, and the inclusion of AI in medical education, at educational levels, as an innovative element of different environments, settings, and learning objectives.

AI in education is a new cycle of technological innovation that, with the development of science and technology, is here to stay. AI is of vital importance in understanding, management, and teacher training for the teaching of new generations for the immediate future that is not long in coming.

Among the techniques that can be identified are the following: Bibliometrics 30% of publications, documentary review 10%, modeling approach, literature review, system description, system evaluation or learning theories 10%, integral effort learning control scheme based on the Barrier-Lyapunov function 10%, problem formulation 10%, statistical expertise domain, machine learning, and robotics 10%, structural equations, virtual reality simulation and machine learning to evaluate surgical experience 10%.

Concerning the selected samples, it can be said that 40% defined it as only empowering reflection articles, another 40% used a sample of articles, 10% worked on a review of documents in companies, and only 10% did it with subjects. This shows an incipient lack of concrete practice in the different social realities. Once the previous results have been determined, it is necessary to define the evolution of AI in education to arrive at the present. For this purpose, the trajectory of scientific articles in the Scopus database was consulted (Figure 3).

AI in education has experienced constant advance since its conception in 1956, establishing its first theoretical antecedents in 1976. Some important milestones refer to the fact that it was in 1956 when the term AI was first discussed at a conference at Dartmouth. In the 1960s-70s, the first AI systems were developed, including pattern recognition and natural language understanding. In the 1980s-90s, advances in information processing and data storage were developed that enabled AI systems to improve on tasks such as machine learning and planning. In the 2000s-2010s, significant advances in deep learning and massive data analysis occurred, allowing AI to improve in tasks such as image recognition and text generation.

From 2015 to the present, AI has continued to evolve at a rapid pace, with advances in technologies such as reinforcement learning, natural language processing, and robotics. AI has also been integrated into a wide variety of practical applications, including facial recognition, virtual assistant, and enterprise data analytics. AI is beginning to be used to personalize education, adjusting the pace and content according to the needs of individual students.

In 2017, AI tools were developed that allowed teachers to evaluate and improve their teaching, as well as allowing students to get real-time feedback on their performance. In 2019, virtual reality and augmented reality began to be used to enhance the learning experience, allowing students to interact with content in a more immersive way.

The COVID-19 pandemic empowered distance education and AI-developing solutions to support online learning becoming a necessity. In 2021, AI was increasingly used to create chatbots and virtual assistants to help students resolve queries and enhance their learning experience and its elements that are still being developed to this day.