Digital

ChatGPT, a prominent large language model developed by OpenAI, has rapidly become embedded in digital infrastructures across various sectors. This narrative review examines its evolving role and societal implications in three key domains: healthcare, education, and the economy. Drawing on recent literature and examples, the review explores ChatGPT’s applications, limitations, and ethical challenges in each context. In healthcare, the model is used to support patient communication and mental health services, while raising concerns about misinformation and privacy. In education, it offers new forms of personalized learning and feedback, but also complicates assessment and equity. In the economy, ChatGPT augments business operations and knowledge work, yet introduces risks related to job displacement, data governance, and automation bias. The review synthesizes these developments to highlight how ChatGPT is driving digital transformation while generating new demands for oversight, regulation, and critical inquiry. It concludes by outlining priorities for future research and policy, emphasizing the need for interdisciplinary collaboration, transparency, and inclusive access as generative AI continues to evolve. Full article

Nutrient analysis through mobile health applications can improve dietary choices, particularly among vulnerable populations. Current mobile nutrient analysis applications face critical limitations: sophisticated deep learning models require substantial computational resources unsuitable for budget devices, while lightweight solutions sacrifice accuracy and lack interpretability necessary for user trust. We develop a lightweight interpretable deep learning architecture combining depthwise separable convolutions, Shuffle Attention mechanisms, and knowledge distillation with integrated Grad-CAM and LIME explanations for real-time interpretability. Our model achieves 97.1% food recognition accuracy (98.0% with cross-validation) and 7.2% mean absolute error in nutrient estimation while maintaining an 11 MB footprint and 150 ms inference time. Knowledge distillation reduces the model size by 62% and energy consumption by 36% while improving accuracy by 2.2 percentage points over non-distilled training. Targeted optimisation for food security categories achieves 94.1% accuracy for staple foods, 93.2% for affordable proteins, and 92.8% for accessible produce. Interpretability methods demonstrate 0.91 feature consistency scores with 38–45 ms explanation generation. These results demonstrate the first mobile nutrient analysis system combining state-of-the-art accuracy with computational efficiency suitable for resource-constrained deployment, addressing accessibility barriers for vulnerable populations. Full article

Many researchers focus on improving reproductive health in sows and ensuring successful breeding by accurately identifying the optimal time of ovulation through estrus detection. One promising non-contact technique involves using computer vision to analyze temperature variations in thermal images of the sow’s vulva. However, variations in camera distance during dataset collection can significantly affect the accuracy of this method, as different distances alter the resolution of the region of interest, causing pixel intensity values to represent varying areas and temperatures. This inconsistency hinders the detection of the subtle temperature differences required to distinguish between estrus and non-estrus states. Moreover, failure to maintain a consistent camera distance, along with external factors such as atmospheric conditions and improper calibration, can distort temperature readings, further compromising data accuracy and reliability. Furthermore, without addressing distance variations, the model’s generalizability diminishes, increasing the likelihood of false positives and negatives and ultimately reducing the effectiveness of estrus detection. In our previously proposed methodology for estrus detection in sows, we utilized YOLOv8 for segmentation and keypoint detection, while monocular depth estimation was used for camera calibration. This calibration helps establish a functional relationship between the measurements in the image (such as distances between labia, the clitoris-to-perineum distance, and vulva perimeter) and the depth distance to the camera, enabling accurate adjustments and calibration for our analysis. Estrus classification is performed by comparing new data points with reference datasets using a three-nearest-neighbor voting system. In this paper, we aim to enhance our previous method by incorporating the mean pixel intensity of the region of interest as an additional factor. We propose a detailed four-step methodology coupled with two stages of evaluation. First, we carefully annotate masks around the vulva to calculate its perimeter precisely. Leveraging the advantages of deep learning, we train a model on these annotated images, enabling segmentation using the cutting-edge YOLOv9 algorithm. This segmentation enables the detection of the sow’s vulva, allowing for analysis of its shape and facilitating the calculation of the mean pixel intensity in the region. Crucially, we use monocular depth estimation from the previous method, establishing a functional link between pixel intensity and the distance to the camera, ensuring accuracy in our analysis. We then introduce a classification approach that differentiates between estrus and non-estrus regions based on the mean pixel intensity of the vulva. This classification method involves calculating Euclidean distances between new data points and reference points from two datasets: one for “estrus” and the other for “non-estrus”. The classification process identifies the five closest neighbors from the datasets and applies a majority voting system to determine the label. A new point is classified as “estrus” if the majority of its nearest neighbors are labeled as estrus; otherwise, it is classified as “non-estrus”. This automated approach offers a robust solution for accurate estrus detection. To validate our method, we propose two evaluation stages: first, a quantitative analysis comparing the performance of our new YOLOv9 segmentation model with the older U-Net and YOLOv8 models. Secondly, we assess the classification process by defining a confusion matrix and comparing the results of our previous method, which used the three nearest points, with those of our new model that utilizes five nearest points. This comparison allows us to evaluate the improvements in accuracy and performance achieved with the updated model. The automation of this vital process holds the potential to revolutionize reproductive health management in agriculture, boosting breeding success rates. Through thorough evaluation and experimentation, our research highlights the transformative power of computer vision, pushing forward more advanced practices in the field. Full article

Cosmetics companies are increasingly integrating augmented reality and artificial intelligence technologies into products and services referred to as beauty tech; consumer perceptions of these solutions, however, remain understudied. Data generated via an online survey are implemented in a stimulus–organism–response framework, deduced from the beauty tech literature. Thereupon, the study identifies how interactivity, informativeness, personalization, and service quality of digital and physical beauty tech solutions for home use affect utilitarian and hedonistic values and the perceived risk factors among consumers. Via customer satisfaction, the effect of the value perception on the purchase intention and loyalty is considered. Results hint at strong effects of characteristics of the services and applications on the utilitarian and the hedonistic dimension of customer experience, which in turn strongly influence customer satisfaction. Perceived risk factors play only a marginal role. Only regarding the tested physical product does higher service quality add to the customer experience. Customer satisfaction in turn results in positive brand perception across different stages of the customer journey and leads to a higher purchase intention, positive brand advocacy, and a higher re-purchase intention. Consequently, well-designed solutions can generate higher customer satisfaction and loyalty on multiple stages along the customer journey. Full article

This research analyzes the digital readiness of Latin American countries by assessing the following key factors: digital infrastructure, human capital, internet use, adoption of digital technology by businesses, and digital government services. These factors are critical to the development of digital technology in the region. The analysis identifies countries that are leaders in digital development (Brazil, Mexico, Chile, Colombia, and Argentina), countries with an average level of digital technology development (Peru, Uruguay, Costa Rica, Paraguay, Panama, and the Dominican Republic), and those with slower progress (Bolivia, Ecuador, Venezuela, Guatemala, El Salvador, Honduras, Cuba, and Nicaragua). Based on this assessment, the study proposes and evaluates positive, negative, and neutral scenarios for the future of digital technology in Latin America over the next five years. The study concludes that a neutral scenario is the most likely, suggesting that, while advanced countries will maintain stable growth, lagging countries will experience accelerated, albeit still moderate, digitalization. This has key implications for regional competitiveness and digital inclusion. The study used methods of analysis, synthesis, classification, grouping, statistics, indexing, and scoring. This study uses the most recent data available (2022–2024) to provide an updated and comprehensive assessment of digital transformation in Latin America, reflecting post-pandemic dynamics and emerging digital trends such as AI and fintech growth. Full article

Increased waste volume and limitations of traditional separation methods have made waste management a hot topic in recent years. To enable the recycling process to be optimized and to minimize environmental impact, waste materials must be well detected and classified. Building on this research, the system is an automated waste-detecting system that integrates machine vision and artificial intelligence (AI). It is coupled with advanced convolutional neural networks (CNNs), which are used for data collection, real-time waste detection, and classification of the proposed framework. Images of waste were captured in many different settings and analyzed with a YOLOv12-based model. The system achieves more gain in detecting and categorizing waste types with 73% precision and a mean average precision (mAP) of 78% in 100 epochs. Results indicate that the YOLOv12 model surpasses the current detection algorithms to provide an efficient and scalable solution to waste management challenges. Full article

Ontology development is a multidisciplinary work involving domain experts and knowledge engineers. Bringing together such a team to develop an ontology of quality is not easy. Therefore, ontologies are often created with limited expertise either in the medical domain or in ontology engineering. Unfortunately, the existing methodologies do not provide much guidance on how the different steps of ontology development should be performed, particularly in the case of reduced involvement of domain experts. This challenge is getting more difficult when there is a multitude of medical knowledge sources and ontologies covering parts of the domain, and often, each has a different representation of the same concept, for example, as a symptom, disease, or clinical sign. This research presents a methodology for creating a medical ontology of quality with limited involvement of the domain experts. The latter are only consulted in the domain definition and evaluation phases. We combine building an ontology from codified knowledge and ontology reuse to enhance reusability and interoperability. The methodology is inspired by METHONTOLOGY, for which we make several improvements, especially in the ontology reuse phase. We provide proof of concept of the proposed methodology with a case study involving the development of the pneumonia diagnosis ontology (PNADO). Full article

The increasing complexity of academic programs and student needs necessitates personalized, data-driven academic advising. Traditional heuristic-based methods often fail to optimize course selection, leading to inefficient academic planning and delayed graduations. This study introduces a hierarchical multi-model machine learning framework for personalized course recommendations, integrating five predictive models: Success Probability Model (SPM), Course Fit Score Model (CFSM), Prerequisite Fulfillment Model (PFM), Graduation Priority Model (GPM), and Recommended Load Model (RLM). These models operate independently in a local model framework, generating specialized predictions that are synthesized by a global model framework through a meta-function. The meta-function aggregates predictions to compute a final score for each course and ensures recommendations align with student success probabilities, program requirements, and workload constraints. It enforces key constraints, such as prerequisite satisfaction, workload optimization, and program-specific requirements, refining recommendations to be both academically viable and institutionally compliant. The framework demonstrated strong predictive performance, with root mean squared error values of 0.00956, 0.011713, and 0.005406 for SPM, CFSM, and RLM, respectively. Classification models for PFM and GPM also yielded high accuracy, exceeding 99%. Designed for modularity and adaptability, the framework allows for the integration of additional predictive models and fine-tuning of recommendation priorities to suit institutional needs. This scalable solution enhances academic advising efficiency by transforming granular model predictions into personalized, actionable course recommendations, supporting students in making informed academic decisions. Full article

The rapid increase in educational data from diverse sources such as learning management systems and assessment records necessitates the application of advanced analytical techniques to identify at-risk students and address persistent issues like dropout rates and academic underperformance. However, many existing models struggle with generalizability and fail to effectively manage data challenges such as class imbalance and missing data, leading to suboptimal predictive performance. This study proposes a hybrid deep learning model combining convolutional neural networks (CNN) and long short-term memory (LSTM) networks to improve the accuracy of student academic performance prediction and enable timely educational interventions. To improve the performance of the model, we incorporate feature selection techniques and optimization strategies to enhance reliability. We also address common preprocessing challenges such as missing data and data imbalance. The proposed model was evaluated on two benchmark datasets to ensure model generalization capability. The hybrid model achieved predictive accuracies of 98.93% and 98.82% on the two datasets, respectively, outperforming traditional machine learning models and standalone deep learning approaches across key performance metrics including accuracy, precision, recall, and F-score. Full article

This study explores the thematic structure and editorial focus of the digital media landscape in Bulgaria by analyzing one year of online news publications from eight major media outlets. The data were collected through a custom-built web scraping application developed in Java, which enabled the automated extraction and processing of full-text articles from publicly accessible news websites. The structured dataset, generated during the scraping process, records word-level occurrences in both article titles and bodies, along with publication dates and URLs. By applying lexical frequency analysis and temporal tracking, this study identifies the most frequently used words and platform-specific usage patterns. The findings reveal clear distinctions in editorial focus between public broadcasters, private national media, and international outlets. Additionally, the analysis highlights how title construction and word prominence vary depending on platform type and media strategy. This study demonstrates the potential of web scraping and computational text analysis as scalable tools for investigating media systems in small and transitional democracies. Full article

Digital trust is increasingly crucial for successful interactions in modern digital environments. However, the existing literature lacks a unified definition and a comprehensive understanding of its core factors. This study addresses these gaps by conducting a systematic literature review to explore and synthesise existing definitions of digital trust and identify the fundamental factors that shape it. A total of 86 relevant sources were analysed, revealing that digital trust is typically conceptualised as confidence in people, processes, and technology aimed at ensuring a secure digital environment, with data protection and privacy playing critical roles. Through thematic analysis, “openness” emerged as an additional factor complementing previously established elements of the integrative model of organisational trust, such as ability, benevolence, and integrity. Based on 42 definitions, we developed a new holistic definition of digital trust. The authors evaluated its content validity, confirming its alignment with the essential factors shaping digital trust’s essence. The findings highlight the multidimensional nature of digital trust and offer an operationalised framework for future measurement and application. Full article

This pilot study explores the efficacy of an AI-powered gamified chatbot integrated with an Online-Merge-Offline (OMO) strategy to enhance parental nutrition knowledge. Conducted in a Taiwanese public childcare setting, the intervention comprised eight weekly nutrition seminars delivered by registered dietitians, supplemented by a LINE-based chatbot providing interactive, gamified learning experiences. Pre-test and post-test evaluations were administered via the chatbot to assess knowledge acquisition. The results from 20 unique participants, including 9 with complete data, indicated a statistically significant improvement in nutritional knowledge (p < 0.0001, Cohen’s d = 2.50), suggesting a substantial educational impact. The integration of gamification elements—such as level completion, community rankings, and personalized feedback—with OMO modalities allowed for sustained engagement, knowledge reinforcement, and seamless transition between digital learning and physical application. This study provides empirical evidence supporting the feasibility and pedagogical value of OMO-gamified chatbots in health promotion and lays the groundwork for future large-scale, longitudinal investigations. Full article

The advent of transformer technology and large language models (LLMs) has further broadened the already extensive application field of artificial intelligence (AI). A large portion of medical records is stored in text format, such as clinical trial texts. Part of these texts is information regarding eligibility criteria. We aimed to harness the immense capabilities of an LLM by fine-tuning an open-source LLM (Llama-2) to develop a classifier from the clinical trial data. We were interested in investigating whether a fine-tuned LLM could better decide the eligibility criteria from the clinical trial text and compare the results with a more traditional method. Such an investigation can help us determine the extent to which we can rely on text-based applications developed from large language models and possibly open new avenues of application in the medical domain. Our results are comparable to the best-performing methods for this task. Since we used state-of-the-art technology, this research has the potential to open new avenues in the field of LLM application in the healthcare sector. Full article

Students with disabilities often face challenges in participating in classroom activities with normal students. Assistive technologies powered by Artificial Intelligence (AI) or Machine Learning (ML) can provide vital support to ensure inclusive and equitable learning environments. In this paper, we identify AI or ML-powered inclusive education tools and technologies, explore the factors required for developing personalized learning plans using AI, and propose a real-time personalized learning framework. We have identified inclusive education tools and technology driven by AI or ML as well as factors impacting the creation of AI-based personalized learning based on our exploration of Google Database, blog sites, company sites, tools, and techniques used in different centers. This study proposes a system model that includes engagement and adaptive learning components. The system uses Bloom’s taxonomy to continuously track the learner’s development. We identified a comprehensive list of AI- or ML-powered inclusive education tools and technologies and determined key factors for developing personalized learning plans, including emotional state, student progress, preferences, learning styles, and outcomes. Based on this research, AI-based inclusive education has the potential to improve educational experiences for students with disabilities by creating a more equitable and inclusive learning environment. Full article

Background: As digital health resources become increasingly prevalent, assessing the quality of information provided by publicly available AI tools is vital for evidence-based patient education. Objective: This study evaluates the accuracy and readability of responses from four large language models—ChatGPT 4.0, ChatGPT 3.5, Google Bard, and Microsoft Bing—in providing contraceptive counseling. Methods: A cross-sectional analysis was conducted using standardized contraception questions, established readability indices, and a panel of blinded OB/GYN physician reviewers comparing model responses to an AAFP benchmark. Results: The models varied in readability and evidence adherence; notably, ChatGPT 3.5 provided more evidence-based responses than GPT-4.0, although all outputs exceeded the recommended 6th-grade reading level. Conclusions: Our findings underscore the need for the further refinement of LLMs to balance clinical accuracy with patient-friendly language, supporting their role as a supplement to clinician counseling. Full article

Mobile Health (mHealth) technologies are transforming healthcare by making it more accessible, efficient, and patient-centric. This study investigates the factors influencing Millennial’s mobile health adoption intention (mHAI). We propose a research model based on the integrated model of the Unified Theory of Acceptance and Use of Technology—UTAUT and the health belief model—HBM. A cross-sectional study was carried out employing purposive sampling to enlist Generation Y (born between 1981 and 1996) and 220 valid questionnaires were collected. We employed structure equation modeling partial least square (SEM-PLS) along with artificial neural network (ANN) and importance–performance map analysis (IPMA) to analyze our model. The research findings revealed that performance expectancy is the most influential factor, while effort expectancy showed no significant association with mHAI. Theoretical and managerial implications are offered to expand the literature on digital healthcare studies, indicating how healthcare providers in developing countries can attract their potential users. Full article

Workplace safety training is the cornerstone of workplace safety and accident prevention. In the case of frequently rotating employees, such as in the laboratories of higher education institutions, where students are required to perform tasks as part of their education, a considerable amount of effort for workplace safety training is demanded from the supervising instructors. Accordingly, the use of self-guided workplace safety training may lead to significant savings in the workload for instructors. In this evaluation study, we investigated to what extent an augmented reality (AR) app is deemed suitable for workplace safety training. The prototypical augmented reality (AR) app is based on an AR platform that performs tracking based on visual Simultaneous Localization and Mapping (vSLAM, Google Tango). The workplace safety training was carried out for two common stations and two devices in a workshop of an environmental engineering laboratory at a higher education institution. A total of 12 participants took part in the mixed-method study. Standardized questionnaires were used to assess usability, cognitive load and the learner prerequisites of motivation and emotion. Qualitative results were collected through subsequent semi-structured interviews. The app was able to achieve good usability, and the values for cognitive load can be classified as conducive to learning, as can the values for the learning prerequisites of motivation and emotion. The interviews provided insights into strengths, but also into potential improvements to the app. The study proved that using vSLAM-based AR apps for workplace safety training is a viable approach. However, for further efforts to establish AR-app-based workplace safety training, these insights need to be ported to a new AR platform, as the platform used has since been discontinued. Full article

The success of digital transformation initiatives relies heavily on effective digital leadership, which requires a blend of human-centric traits and technical expertise. While digital technologies enable transformation, organizations must develop leaders with the skills to navigate the complexities of change, foster innovation, and align strategies with organizational goals. Despite the growing importance of digital leadership, there is a lack of standardized, validated tools to measure and assess digital leadership competencies systematically. This study introduces the Digital Leadership Scale (DLS), a validated self-assessment tool designed to measure a leader’s ability across seven human-centric dimensions essential for digital transformation: Positive Attitude, Ethical AI Use, Growth Mindset, Track Record, Transparent Agenda, Skills Acquisition, and Participative Style. The DLS serves as a practical tool for leaders to engage in self-reflection, identify strengths and development areas, and adopt personalized learning strategies. Organizations can leverage this scale to cultivate a digitally proficient workforce and foster leadership capabilities aligned with digital transformation success. Full article

This paper presents an application of the Ant Colony Optimization (ACO) algorithm combined with the Logistic Regression (LR) method in the lead acid battery charging process. The ACO algorithm is used to obtain the best current pattern in the battery charging system to produce a smart charging system with a fast and safe charging current for the battery. The best current pattern is conducted gradually and repeatedly to obtain termination in the form of the best current pattern according to the ACO algorithm. The results of the algorithm design produce a current pattern consisting of 10 A, 5 A, 3 A, 2 A, and 0 A. The charging system with this algorithm can charge all types of lead acid batteries. In this research, the capacity of battery 1’s State of Charge (SOC) is 56%, battery 2’s SOC is 62%, and battery 3’s SOC is 80%. When recharging the battery’s full condition to a SOC of 100%, the length of time for charging battery 1 for 12.73 min, battery 2 takes 15.73 min, and battery 3 takes 29.11 min. Smart charging with the ACO can charge the battery safely without current fluctuations compared to charging without an algorithm such that the amount of charging current used is not dangerous for the battery. In addition, data analysis is carried out to determine the value of accuracy in estimating SOC charging using supervised learning linear regression. The results of the data analysis with linear regression show that the battery’s SOC estimation has good accuracy, with an RMSE value of 0.32 and an MAE of 0.27. Full article