The reliability of information systems infrastructure is critical. After all, it is the basis of other business systems and processes in the organization. In a turbulent world development of a robust and reliable information system infrastructure is a challenge. Infrastructure should run as a solid foundation for operations, but be flexible when needs change. Information systems infrastructure should also provide a shield against different kinds of threats, such as viruses and hacking from the outside world. Understanding elements that impact the reliability of infrastructure is therefore an important element in the development of reliable information systems infrastructure.
The rapid expansion of generative AI (GenAI) in higher education raises important questions about what drives its sustained and responsible use. This study applies an extended UTAUT2 framework to examine GenAI adoption among social science students and researchers. Using PLS SEM on responses from 569 participants across Baltic Sea region universities, the study evaluates how performance expectancy, effort expectancy, social norms, study value, facilitating conditions, hedonic motivation, and habit shape self-reported GenAI use in classroom tasks, homework, self-study, and thesis work. Habit emerges as the strongest predictor for both groups (students = 0.420; staff = 0.712), explaining substantial variance (R² = 0.479 and 0.558). Students exhibit modest positive effects from performance expectancy, effort expectancy, social norms, and study value. In contrast, study value is the sole additional predictor identified for staff. These findings underscore the importance of scaffolded GenAI activities that foster responsible habits, as well as institutional strategies that prioritize demonstrable academic value.
Digital occupational safety and health training is nowadays widely adopted in complex socio-technical systems. Yet its effectiveness is often assessed using linear approaches that fail to capture impacts on real work and system performance. This study argues that training effectiveness cannot be meaningfully evaluated without a systemic perspective. To address this gap, the paper proposes the FRAM-Kirkpatrick Framework, integrating the Functional Resonance Analysis Method (FRAM) with the Kirkpatrick training evaluation model. The framework conceptualizes digital OSH training as a systemic intervention influencing functional performance variability. The FRAM-Kirkpatrick Framework is applied through multiple case studies in the road transport sector involving three companies and 47 professional drivers. Results show significant improvements in learning and safe behavior, accompanied by a substantial reduction in functional variability within critical system functions. The findings demonstrate that the FRAM-Kirkpatrick Framework enables a more robust evaluation of digital training effectiveness in complex socio-technical systems.
This study looks at how the data-based approach to group formation can improve student behavior in the classroom. We built a system called the Student Profile Vector (SPV) to capture the emotional details of each student. We used cosine similarity and Genetic Algorithms to put the students into the groups that share traits. We tested the model in five institutions. We found that the model does not just reduce the classroom conflict—it increases the student engagement a lot. Our findings show that the computational tools are practical and scalable. The computational tools can make the learning environment more inclusive and adaptive. We see this as an approach.
To facilitate teachers convey educational scenarios, the study reviews five learning design tools: Leaarning Designer, eXe Learning, LAMS, EdCrumble, and Cadmos. Previous research emphasizes the importance of learning design in enhancing teaching quality and simplifying the selection of appropriate tools to cater to educators' varying levels of experience, especially new teachers. The primary aim is to assess the ability of these tools to support teachers throughout the three design phases: conceptual, authoring, and implementation/sharing enabling them to effectively capture and represent a two-hour digital literacy scenario. Forty teachers from a Vocational High School participated in the research. They attended workshops where each design tool was used to model an educational scenario aligned with the curriculum. Workshops introduced teachers to each tool, followed by a practical application of representing a predefined scenario. Data were collected through questionnaires, which were analyzed for reliability and trends in satisfaction. Statistical indicators, such as Cronbach’s alpha, were used to validate the results. The tools demonstrated varying strengths: Learning Designer excelled in conceptual guidance, while edCrumble offered comprehensive graphical summaries. Graphical representation was strong in LAMS, edCrumble, and Cadmos, whereas Learning Designer and eXe Learning leaned towards verbal descriptions. Key gaps included limited support for metacognitive objectives and resource classification. Implementation in LMS systems and sharing capabilities also varied among the tools. The study concludes that although the tools meaningfully support teachers in learning design, enhancements are necessary to address existing gaps.
Psychological analysis of drawings has long been a subject of interest in clinical and research settings, offering insights into an individual's personality, emotional state, and cognitive processes. Traditional methods rely on subjective interpretation, often leading to bias and inconsistency. This study presents a novel approach to psychological drawing analysis by integrating artificial intelligence and computer vision techniques. A Python-based application has been developed, utilizing machine learning algorithms for automated recognition of geometric shapes, color patterns, and symmetry within drawings. The system employs convolutional neural networks for shape detection and a structured color recognition model based on the RGB model. The extracted visual features are analyzed using a generative artificial intelligence model to infer psychological traits and emotional states. While the application does not replace clinical diagnosis, it offers a valuable complementary tool for therapists by providing objective and data-driven insights. The research was designed as a proof-of-concept study, aiming to demonstrate the technical feasibility and potential clinical relevance of AI-assisted drawing analysis. Future work will focus on expanding the dataset, applying statistical validation, and improving accuracy, interpretive depth, and accessibility via mobile and cloud platforms.
Digital educational escape rooms have emerged over the past decade as an innovative, game-based learning approach. Although interest in escape rooms is steadily increasing, the literature shows a lack of a widely accepted methodological framework for their design and implementation in educational settings. In this context, the present study integrates Project-Based Learning with the STEAM approach through the use of a well-designed educational scenario called ‘STEAMscapers’. This scenario is implemented within an e-learning environment, using Web 2.0 technologies and the escape room format to assess the effectiveness of a pedagogically designed solution hosted on the WIX platform. The innovative aspect of this research is that STEAMscapers represents the first fully developed instructional scenario that combines both STEAM and Project-Based Learning within a digital escape room environment. To assess its effectiveness, a quantitative study was conducted with a sample of 40 e-learning experts. The research focused on six key effectiveness dimensions, referred to as Key Performance Indicators (KPIs). The results were highly positive, with participants rating the escape room as very effective—particularly in terms of content quality and interactivity—and also giving high scores in other areas, such as assessment, learning outcomes and course organization. The study concludes that digital escape rooms, combining Project-Based Learning with the STEAM approach, can be effective educational tools for primary education.
This article aims to investigate whether the use of the internet and social networks contributes to or influences digital skills in Romania. The digitalization gap in Romania is calculated as the difference between internet access/use and digital skills, and has been compared with that of the European Union. A descriptive and comparative approach was used to analyze data from the European Commission, Statista, Datareportal, and Eurostat. This research sought to identify structural patterns between internet access, internet use, social media, and digital skills, and did not aim to present causal inferences. A quantitative approach was also applied in this study. As such, to reduce this digital gap and also bring innovation to Romania's economy, the article presents solutions in the field of artificial intelligence tools. The article's conclusions indicate that integrating tools based on AI and learning assistants supports the learning process and can reduce the digitalization gap.
This document describes the role of Information Systems Analysis and Design (ISAD) and Design Thinking (DT) in the development of a computerized simulator for direct current machines within the Electrical Machines Laboratory at the National University of Colombia, Bogotá campus. A brief description of both methodologies, the developed simulator, and the physical environment of the Laboratory where it is implemented is provided. Design Thinking (DT) and Information Systems Analysis and Design (ISAD) are complementary methodologies that have proven effective in developing technological solutions, particularly in industrial automation. This paper presents the application of these methodologies in the simulator’s development. The simulator not only replicates basic operating conditions and wiring of involved components but also integrates advanced automatic control techniques, simulation, and uncertainty quantification methods for enhanced reliability assessment.
This comprehensive research examines the impact of Artificial Intelligence (AI) on industrial and mechanical engineering education by synthesizing insights from 12 review papers published between 2023 and 2025. The study employed CiteSpace software for co-citation and keyword co-occurrence analysis to identify key intellectual structures and thematic clusters shaping the field. The co-citation analysis of authors highlights three main research areas: the use of generative AI tools (e.g., ChatGPT) for personalized education; data-driven models for predicting academic success; and, not least, immersive technologies such as robotics and virtual reality for experiential learning. The clustering of co-cited literature reveals two primary domains: (1) practical AI integration aimed at improving curriculum and exam assessments, and (2) the future potential of generative and immersive AI technologies to promote creativity and skill development among next-generation professionals. The co-citation networks of authors emphasize two distinct but overlapping scholarly communities: one focused on system-level educational technologies and the other on ethical, curricular, and institutional reform needs. The keyword co-occurrence patterns indicate convergence around two main themes: on the one hand, AI-powered analytics and simulations, while on the other, learner-centered adaptive systems. Overall, the findings of this umbrella study suggest a paradigm shift toward AI-enhanced, student-centered, and policy-driven engineering education, as this interdisciplinary field rapidly develops and widens its application, with critical societal relevance for the future.
This paper proposes a method for generating and recommending personalized evacuation routes in cases of natural disasters or industrial accidents—a topic of significant importance given the recent increase in such events. The utilization of the method is demonstrated on an example of a disaster/accident, having an epicenter, which is spreading radially in all directions with the same speed of onset. The web-based app, specially developed for this purpose, is briefly presented as well.
This umbrella mapping review examines the integration of Artificial Intelligence (AI) into accounting education by synthesizing recent review-type studies. A PRISMA-guided search of Web of Science identified the corpus, while bibliometric science mapping with CiteSpace was used to uncover underlying structures and trends. To improve topical clarity and analytical accuracy, this paper proposes a Methodological Framework for Key Terms Mapping (MFKTM), which aims to systematically extract and categorize key terms across multiple sections of each article, supporting more precise and context-aware thematic analysis. Through co-citation and co-occurrence analyses of references, authors, and keywords, the study identifies five interconnected thematic clusters: curriculum reform, technological adoption, competency development, ethical implications, and systemic transformation. The co-citation network of authors highlights key intellectual influencers, while the temporal evolution of keywords reveals shifting research priorities toward acceleration and convergence around AI-supported assessment and AI-literacy outcomes. Overall, the findings illustrate a multi-level reconfiguration of accounting education, positioning AI not just as a technological enhancement but as a catalyst for broader pedagogical and institutional change. This study provides a transdisciplinary roadmap for aligning academic curricula with the evolving demands of the digital economy.
AI is a key tool revolutionizing today's society, market, and economy. Its applications are everywhere. In such an environment, investigating the attitudes of Hellenic Naval Academy (HNA) students towards AI is a challenge, particularly concerning its acceptance in the military sector and education. This study used a tool (questionnaire) to investigate attitudes and revealed important findings by using the pre-test & post-test procedure.
large growth of e-sports has defined it as an important cultural and economic phenomenon. Millions compete with each other, while millions view them worldwide. Despite the importance of this topic, educational materials about e-sports history and culture are still scarce. Therefore, this study presents the development of EsportLingo as an innovative gamified educational platform designed with the goal of bridging this gap. This application is a progressive web application, and it offers accessibility, interactivity, as well as scalability. It includes theoretical modules, quizzes, and leaderboards to increase user engagement and help learning. The results show that EsportLingo has the potential to combine gamification and education to help the e-sports community and inform the next generation of enthusiasts.
This paper presents a review of the applications of Artificial Intelligence (AI) in education. This review aims to explore how artificial intelligence (AI) is being used in education and how it can improve both the learning process and the results for students. Moreover, the purpose of this research is to investigate how AI bolsters contemporary pedagogical practices that captivate and inspire students. A qualitative research design using a retrogressive approach was adopted to analyze previous studies and literature. As a result, four key AI applications in education were identified: profiling and prediction, Intelligent Tutoring Systems, automation, and educational robots and chatbots. Those AI applications enrich the learning process and improve the learning results. In conclusion, this research has improved the significance that AI can have in education by offering more adaptive, responsive, and personalized experiences, as well as supporting the shift to modern teaching methods.
has a significant impact on learning effectiveness, making it crucial to study attention. However, there is little research on the quantitative measurement of learning attention. Electroencephalography (EEG) signals can reflect the brain's attention during learning; therefore, this paper proposes a learning concentration detection method based on EEG. Firstly, a portable, wearable single-channel EEG acquisition device is used to collect the brain's EEG signals during the learning process. Secondly, the single-channel EEG signals are converted into images to evaluate learning concentration, thereby transforming the concentration detection problem into an image recognition task. Thirdly, convolutional neural network models - AlexNet, ResNet, and the Visual Geometry Group (VGG) network-are applied to detect the converted images. Finally, an experiment was conducted, and the results show that the detection accuracy rate reaches 93.23%, which proves that the proposed method can effectively evaluate learning concentration.
This paper introduces the design, development, and deployment of an NFC-based smart card system tailored explicitly for academic environments, aiming to improve data management, student identification, and administrative automation. Built around the Arduino Uno R3 and PN532 NFC module, the system provides seamless integration with a MariaDB relational database and a Java-based user interface. Key features include student attendance tracking, real-time access to academic records, and secure cloud-based data storage. A role-based access model is implemented to ensure that students and professors have appropriate visibility of data, thereby reinforcing data privacy and security. The system enables students to interact with NFC cards using their smartphones, granting access to personalized academic files stored on platforms such as Google Drive. The software layer, developed using Visual Studio Code and Apache POI for Excel exports, enables robust administrative control over student records, grades, and catalog updates.
The study examines the usage of digital platforms and student engagement in secondary schools in urban and rural areas of Northern Saudi Arabia. Empirical data were collected from 30 schools (15 urban, 15 rural) through quantitative surveys and teacher-reported metrics on engagement. The study examines four popular platforms-Madrasati, Google Classroom, Microsoft Teams, and WhatsApp-through three domains of engagement: behavioural, emotional, and cognitive. The results reflect significant differences in the platform's effectiveness, teacher digital fluency, and infrastructure accessibility. Madrasati proved to be the best-performing platform, recording mean cognitive engagement scores of 4.3/5 for urban school learners vis-& agrave;-vis 3.5/5 for rural learners. Google Classroom and Teams proved to be of medium effectiveness, whereas WhatsApp, despite widespread usage in the countryside (75%), recorded the lowest cognitive engagement (2.3/5). Urban learners had significantly higher platform usage overall, with 94% claiming personal device availability and 89% having stable internet connectivity, compared to 57% and 43%, respectively, for rural learners. Teacher preparedness was significantly linked to student participation. Within urban schools, 85% of the teachers were formally trained and registered, with an average 4.3/5 score of digital fluency, while just 41% rural educators were similarly prepared, with an average fluency score of 2.8/5. These gaps demonstrate that technology implementation in isolation cannot be adequate without equal investment in infrastructure and educator development. The research concludes that effective digital learning depends not solely on platform accessibility but equally on teacher capacity and national capability. Policy suggestions highlight targeted capacity building, the localization of platforms, and increased infrastructure to guarantee inclusive digital involvement to align with Vision 2030.
Computational thinking has gained an important place in modern education, enabling individuals to approach problem-solving in a logical and structured manner. This cross-curricular competence is important and applicable in any field of science, not just for computer science professionals. By fostering problem-solving, critical thinking, and creativity, among other skills, computational thinking is crucial in today's education. In the digital age, computational thinking is not just a technical skill, or one related to programming and robotics, but a way of thinking that can transform, or at least provide a different perspective, the way we approach everyday challenges and opportunities in our daily lives. To assess this new competency, analytical tools and methods that are not too general are needed. To achieve this, that is, to assess computational thinking, the process is currently complex and requires a combination of qualitative and quantitative methods. In this way, analytical rubrics, portfolio analysis, and standardized tests are essential tools that help provide a comprehensive and accurate assessment of students' skills related to this competence. In our project, we also work on assessing computational thinking using Bebras-type tasks and applying data analysis. Data analysis facilitates the continuous improvement of teaching and assessment methods. By monitoring and analyzing data over time, educators can identify the most effective strategies and make adjustments to improve learning outcomes. In this paper, we introduce COMATH, an assessment tool grounded in research, which has undergone two phases of piloting across six counties. This process involved collaboration with subject-matter experts and the participation of over 4500 students and 100 teachers. We employ tasks designed to evaluate computational thinking and share some of the findings we have gathered to date.
This study aims to develop a predictive model for student success by integrating machine learning algorithms with learning style analysis. Educational institutions increasingly recognize the value of early performance prediction to implement timely interventions and enhance learning outcomes. Learning management systems generate vast amounts of data. The proposed research will analyze student interaction data from Moodle learning management system, including course logins, resource access patterns, assignment submissions, and assessment performance. These digital footprints will be combined with learning style assessments to identify patterns of academic achievement. Machine learning algorithms are applied and compared to determine the most effective predictive model. This research contributes to educational data mining by exploring the intersection between digital behavior patterns, individual learning preferences, and academic outcomes. The resulting model achieves high prediction accuracy, enabling proactive educational interventions that adapt to students' learning styles while leveraging Moodle's AI capabilities for personalized learning experiences.
