Artificial intelligence (AI) has rapidly evolved, transforming industries and addressing societal challenges across sectors such as healthcare and education. This study provides a state-of-the-art overview of AI research up to 2023 through a bibliometric analysis of the 50 most influential papers, identified using Scopus citation metrics. The selected works, averaging 74 citations each, encompass original research, reviews, and editorials, demonstrating a diversity of impactful contributions. Over 300 contributing authors and significant international collaboration highlight AI’s global and multidisciplinary nature. Our analysis reveals that research is concentrated in core journals, as described by Bradford’s Law, with leading contributions from institutions in the United States, China, Canada, the United Kingdom, and Australia. Trends in authorship underscore the growing role of generative AI systems in advancing knowledge dissemination. The findings illustrate AI’s transformative potential in practical applications, such as enabling early disease detection and precision medicine in healthcare and fostering adaptive learning systems and accessibility in education. By examining the dynamics of collaboration, geographic productivity, and institutional influence, this study sheds light on the innovation drivers shaping the AI field. The results emphasize the need for responsible AI development to maximize societal benefits and mitigate risks. This research provides an evidence-based understanding of AI’s progress and sets the stage for future advancements. It aims to inform stakeholders and contribute to the ongoing scientific discourse, offering insights into AI’s impact at a time of unprecedented global interest and investment.

The integration of Big Earth Data and Artificial Intelligence (AI) has revolutionized geological and mineral mapping by delivering enhanced accuracy, efficiency, and scalability in analyzing large-scale remote sensing datasets. This study appraisals the application of advanced AI techniques, including machine learning and deep learning models such as Convolutional Neural Networks (CNNs), to multispectral and hyperspectral data for the identification and classification of geological formations and mineral deposits. The manuscript provides a critical analysis of AI's capabilities, emphasizing its current significance and potential as demonstrated by organizations like NASA in managing complex geospatial datasets. A detailed examination of selected AI methodologies, criteria for case selection, and ethical and social impacts enriches the discussion, addressing gaps in the responsible application of AI in geosciences. The findings highlight notable improvements in detecting complex spatial patterns and subtle spectral signatures, advancing the generation of precise geological maps. Quantitative analyses compare AI-driven approaches with traditional techniques, underscoring their superiority in performance metrics such as accuracy and computational efficiency. The study also proposes solutions to challenges such as data quality, model transparency, and computational demands. By integrating enhanced visual aids and practical case studies, the research underscores its innovations in algorithmic breakthroughs and geospatial data integration. These contributions advance the growing body of knowledge in Big Earth Data and geosciences, setting a foundation for responsible, equitable, and impactful future applications of AI in geological and mineral mapping.

The advent of Artificial Intelligence (AI) has transformed Learning Management Systems (LMSs), enabled personalized adaptation and facilitated distance education. This study employs a bibliometric analysis based on PRISMA-2020 to examine the integration of AI in LMSs from an educational perspective. Despite the rapid progress observed in this field, the literature reveals gaps in the effectiveness and acceptance of virtual assistants in educational contexts. Therefore, the objective of this study is to examine research trends on the use of AI in LMSs. The results indicate a quadratic polynomial growth of 99.42%, with the years 2021 and 2015 representing the most significant growth. Thematic references include authors such as Li J and Cavus N, the journal Lecture Notes in Computer Science, and countries such as China and India. The thematic evolution can be observed from topics such as regression analysis to LMS and e-learning. The terms e-learning, ontology, and ant colony optimization are highlighted in the thematic clusters. A temporal analysis reveals that suggestions such as a Cartesian plane and a league table offer a detailed view of the evolution of key terms. This analysis reveals that emerging and growing words such as Learning Style and Learning Management Systems are worthy of further investigation. The development of a future research agenda emerges as a key need to address gaps.

This research was conducted using a survey research method to investigate the influence of Artificial Intelligence (AI) on Nigerian students' academic performances in tertiary institutions. Nigerian tertiary institutions have an estimated population of about 2.5 million students across the universities, polytechnics, monotechnics, and colleges of education. A sample size of 509 was used. The researchers adopted an online questionnaire (Google Form) to administer questions to respondents across Nigeria to elicit responses from the respondents bordering on their awareness and the use of AI and its attendant impacts on their academic performance. Five research objectives were raised for the proper investigation of this study. From the findings of the study, the researchers found that the majority of Nigerian students use AI and that AI has positive impacts on the educational performance of Nigerian students. It was also found that Nigerian students have training on the use of AI for educational purposes and that they are more familiar with Snapchat AI and ChatGPT. Conclusively, AI is useful to students in the sense that it enhances their knowledge of their courses, improves their learning and speaking skills, and helps them to have a quick understanding of their course by way of simplifying technical aspects of their courses. The researchers therefore recommend as follows: Nigerian tertiary institutions should formally train students as well as teachers on the use of AI for academic purposes so that they can understand the ethical implications of the use of AI. Using AI for writing could be interpreted to mean examination malpractice, and this should not be condoned in the educational sector; however, at the moment, a small number of students used AI for examinations. Albeit, the appropriate use of AI should be fully integrated into Nigerian tertiary institutions' curricula.

Brain tumors are a primary factor causing cancer-related deaths globally, and their classification remains a significant research challenge due to the variability in tumor intensity, size, and shape, as well as the similar appearances of different tumor types. Accurate differentiation is further complicated by these factors, making diagnosis difficult even with advanced imaging techniques such as magnetic resonance imaging (MRI). Recent techniques in artificial intelligence (AI), in particular deep learning (DL), have improved the speed and accuracy of medical image analysis, but they still face challenges like overfitting and the need for large annotated datasets. This study addresses these challenges by presenting two approaches for brain tumor classification using MRI images. The first approach involves fine-tuning transfer learning cutting-edge models, including SEResNet, ConvNeXtBase, and ResNet101V2, with global average pooling 2D and dropout layers to minimize overfitting and reduce the need for extensive preprocessing. The second approach leverages the Vision Transformer (ViT), optimized with the AdamW optimizer and extensive data augmentation. Experiments on the BT-Large-4C dataset demonstrate that SEResNet achieves the highest accuracy of 97.96%, surpassing ViT’s 95.4%. These results suggest that fine-tuning and transfer learning models are more effective at addressing the challenges of overfitting and dataset limitations, ultimately outperforming the Vision Transformer and existing state-of-the-art techniques in brain tumor classification.