This study examines how artificial intelligence (AI) generates operational value in data centers through the Task–Technology Fit (TTF) framework. Using survey data from data center professionals, the study applies covariance-based structural equation modeling (CB-SEM) to test relationships among task characteristics, technology characteristics, managerial competence, organizational support, TTF, AI adoption, and operational efficiency. The findings indicate that task, technology, and managerial factors strengthen TTF, which serves as the key mechanism driving AI adoption. Organizational support also facilitates adoption by creating conditions that enable effective technology integration. AI adoption, in turn, enhances operational efficiency. The study contributes to AI adoption research by showing that AI value depends not only on technological capability but also on alignment with task demands, managerial expertise, and supportive organizational contexts. It offers practical insights for improving AI implementation in complex data center environments.
As artificial intelligence (AI) continues to shape commercial and governmental innovation, understanding the potential of large language models (LLMs) to improve software testing has become increasingly critical. This paper explores the application of LLMs for generating test cases-an essential yet resource-intensive activity in software quality assurance. Using theoretical lenses of the Design Science Research Evaluation (DSRE) framework, Measurement Theory, and Cognitive Load Theory, the authors systematically design and evaluate procedures for test case generation and execution to benchmark AI-generated test cases against those produced by human testers and traditional pairwise analysis. They develop two custom-built applications seeded with intentional defects and execute test cases generated through each method. This research contributes to the evolving discourse on AI-driven software engineering, offering insights into when LLMs can autonomously generate test cases, when human expertise is indispensable, and when hybrid approaches yield the greatest value.
The COVID-19 pandemic accelerated the adoption of wearable health technologies (WHTs), reshaping patterns of use and motivations. This study applies Social Cognitive Theory and Digital Divide frameworks to compare pre-and post-pandemic determinants of WHT adoption using survey-weighted logistic regressions and subgroup analyses by race and gender. Findings show that while personal agency and social context remain relevant, structural factors such as smart device ownership, insurance coverage, and digital health integration became more decisive after the pandemic. The influence of education declined, and income-based disparities widened, signaling a shift from skill-based divides to access-based divides. Demographic trends remained stable, although health-risk salience increased and equity patterns diverged. The results suggest that motivation and confidence alone are insufficient to influence WHT adoption. Material access and digital infrastructure now play a central role. Implications are discussed for temporally sensitive IS theory and inclusive technology design and policy.
As artificial intelligence (AI) continues to shape commercial and governmental innovation, understanding the potential of large language models (LLMs) to improve software testing has become increasingly critical. This paper explores the application of LLMs for generating test cases—an essential yet resource-intensive activity in software quality assurance. Using theoretical lenses of the Design Science Research Evaluation (DSRE) framework, Measurement Theory, and Cognitive Load Theory, the authors systematically design and evaluate procedures for test case generation and execution to benchmark AI-generated test cases against those produced by human testers and traditional pairwise analysis. They develop two custom-built applications seeded with intentional defects and execute test cases generated through each method. This research contributes to the evolving discourse on AI-driven software engineering, offering insights into when LLMs can autonomously generate test cases, when human expertise is indispensable, and when hybrid approaches yield the greatest value.
In the Web 3.0 era, blockchain emerges as a disruptive technology offering robust solutions for industries aiming to optimize and sustain their supply chains. With features such as decentralization, transparency, traceability, and immutability, blockchain ensures data security, prevents corruption or unauthorized access, and enables seamless interoperability. This study identifies and analyzes 10 key strategies for blockchain adoption in supply chains through a rigorous PRISMA-based literature review. Using the m-TISM methodology, a strategic framework is developed to define interrelationships among strategies, followed by a MICMAC analysis to classify them based on driving power and dependence. Findings highlight three pivotal strategies: identifying complex processes suited for blockchain, spreading awareness among supply chain partners, and fostering training and capacity building. These strategies drive effective adoption across industries. Further, a path analysis suggests three implementation pathways, offering industries a structured roadmap for smooth blockchain integration.
Social media has become integral to daily interactions and a key data source for researchers. Using COVID-19 as a case study, this work compares 24 social media datasets to address three research questions: 1) Is the dataset in compliance with the FAIR principles of being Findable, Accessible, Interoperable, and Reusable? 2) To what extent have people utilized social media to voice and exchange their apprehensions during the COVID-19 pandemic? 3) To what extent can social media datasets be utilized for natural language processing (NLP)-based COVID-19 pandemic surveillance? Leveraging the evaluation questions derived from the FAIR principles, the authors assess 24 social media datasets related to the COVID-19 pandemic. Additionally, they comprehensively analyze each dataset, including their composition, and the specific instances and features they encompass. They have initiated an attempt hoping that more researchers will join to create a data community where information can be repurposed and reused.
Digital agriculture is a growing field where complex systems are utilized for computer vision applications to aid decision-making. This work provides a comprehensive comparison of the applications of Convolutional Neural Networks and Vision Transformers, in the agricultural sector, specifically focusing on image classification. An extensive evaluation and comparison of both modeling frameworks is conducted with 17 different deep learning models. Using five distinct agricultural datasets comprising various image classification tasks, this work highlights the strengths and limitations of each architectural approach. Specifically, transformer variants like the Swin Transformer Version 2 excel in accuracy and the ability to capture complex patterns due to their attention-based mechanisms. In contrast, models Residual Networks offer a balance between computational efficiency and performance, making them suitable for scenarios with limited computational resources. The findings highlight the importance of selecting an appropriate deep learning model based on specific agricultural tasks.
The rapid proliferation of fake news on social media threatens public trust and social stability. The authors propose a Responsible AI (RAI) framework for multimodal fake news detection that integrates threat theory with deep learning. They introduce two responsibility indicators: (1) User Trust Score (UTS), measuring source credibility via trustworthiness, expertise, and behavioral consistency; (2) Responsible News Dissemination Score (RNDS), quantifying content alignment with responsible discourse using a 66-keyword RAI vocabulary. These indicators are integrated into Graph Neural Networks through node-level feature augmentation and responsibility-aware message passing. The model fuses BERT text, VGG19 visual features, propagation patterns, and responsibility metrics, achieving F1-scores of 0.9498 on Politifact and 0.9720 on Gossipcop. They discuss ethical risks including algorithmic stigmatization, privacy concerns, and demographic bias, and propose mitigation strategies such as temporal score decay, federated learning, and fairness audits to ensure responsible deployment.
Objective eye movement data have the potential to measure users' states instantly and in real time, providing a basis for timely intervention and personalized adaptation. Despite the broad applicability of eye-based construct measurement, research on its development and validation remains limited. Following the preferred reporting items for systematic reviews and meta-analyses guidelines, this review analyzes 127 studies that investigate the use of eye-tracking metrics to measure abstract constructs and the corresponding validity evidence. The findings reveal that eye-tracking metrics can measure a wide variety of constructs. Drawing on validity evidence commonly employed in psychometric-based construct measurement, this study synthesizes and summarizes validity evidence applicable to eye-based construct measurement. To illustrate the application of eye-tracking metrics and their supporting evidence, this article uses the example of detecting a vehicle driver's cognitive load, offering guidance for future studies and practical applications.
This article explores digital business models in the Metaverse, a virtual space revolutionizing traditional models. It examines the Metaverse’s immersive nature, its potential to disrupt businesses, and the importance of data-driven value propositions, personalized experiences, digital platforms, agility, disintermediation, and innovation. The research highlights the Metaverse’s impact on digital models, emphasizing adaptation. It discusses challenges and opportunities, focusing on interoperability, decentralization, social interactions, and user-generated content. The article provides insights and practical guidelines for designing Metaverse-specific business models, incorporating immersive experiences, virtual assets, and systematic strategies to leverage its unique features.
Objective eye movement data have the potential to measure users’ states instantly and in real time, providing a basis for timely intervention and personalized adaptation. Despite the broad applicability of eye-based construct measurement, research on its development and validation remains limited. Following the preferred reporting items for systematic reviews and meta-analyses guidelines, this review analyzes 127 studies that investigate the use of eye-tracking metrics to measure abstract constructs and the corresponding validity evidence. The findings reveal that eye-tracking metrics can measure a wide variety of constructs. Drawing on validity evidence commonly employed in psychometric-based construct measurement, this study synthesizes and summarizes validity evidence applicable to eye-based construct measurement. To illustrate the application of eye-tracking metrics and their supporting evidence, this article uses the example of detecting a vehicle driver’s cognitive load, offering guidance for future studies and practical applications.
The convergence of digital twin technology and data analytics continues in the area of smart cities focusing on the comprehensive study of data analysis and its visualization. It begins by standing a foundational framework for data analytics and discussing the importance of these ideas in figuring out complex patterns, concluding, and supporting thoughtful decision-making. The article emphasizes the crucial role of data analytics for urban innovation in the context of smart cities using digital twins. The study delves into the complexities of data collection, integration challenges, and innovative solutions, underscoring the necessity of constructing a robust digital twin ecosystem with a variety of sensors and data sources with its visualization. Smart recommendations by monitoring, prescriptive, and real-time analytics are becoming essential tools for vigilant urban management for taking the best and next course of action. The article delves into predictive analytics, highlighting the synergy of data streams for a comprehensive understanding of urban dynamics.
Protection of networks from changing cyberthreats depends critically on intrusion detection. This article presents a hybrid deep learning framework using a tunicate swarm algorithm and brown-bear optimization for intrusion detection. The Tunicate Swarm Algorithm (TSA) was utilized for hyperparameter tuning; the Brown-Bear Optimization Algorithm (BBOA) was employed for feature selection, therefore lowering the dataset from 41 to 25 features. After five epochs, the model tested on the NSL-KDD dataset achieves 98% accuracy. Comparative study using conventional models showed that the suggested framework improved accuracy and loss reduction, therefore stressing its possibilities to improve intrusion detection systems.
Error detection is an important part of preparing data for data analysis. Erroneous data can result in inaccurate analysis, resulting in garbage-in, garbage-out. Currently, many models utilize either or both Qualitative and Quantitative methods to detect errors in the data. However, these methods are still limited in the errors they can detect. Hence FADE, Focused and Attention-based Detector of Errors, was proposed. FADE can detect errors within structured data with rows and columns. FADE utilizes the information found in surrounding cells within the same row to help determine if a cell is erroneous. It also learns the expected structure of the attributes in the dataset and the values expected in each attribute. This results in FADE having a much wider range of error type detection and having a higher classification of errors than other methods. FADE was evaluated and was found to detect these errors with relatively high performance.
Massive amounts of data, including big data, are generated and collected today from a variety of diverse data sources. These big data differ in terms of their veracity that ranges from imprecise and uncertain to precise. These data hide a huge amount of valuable information and precious knowledge that ought to be discovered. Examples of big data in the healthcare and epidemiological fields include information about patients afflicted with diseases such as Coronavirus disease 2019 (COVID-19). Researchers, epidemiologists, and policy makers get a great deal of help from the knowledge discovered from these data via data science techniques such as machine learning, data mining and online analytical processing (OLAP) in order to fully uncover the secrets of the disease. Eventually that may also inspire them to come up with ways to detect, control and fight the disease. In the article, the authors present a machine learning and big data analytical tool useful to process and analyze COVID-19 epidemiological data, while supporting big data visualization and visual analytics.
This study introduces a new meta-explainable machine learning methodology to enhance medical care recommendations and optimize healthcare operations through targeted interventions. It could assist a large, and diverse population facing challenges in resource allocation and operational complexity. The proposed method utilizes a two-stage model. It first employs an Explainable Boosting Machine (EBM) and then provides the output from the initial phase to an unsupervised machine learning framework. It examines diverse aspects to identify the most critical set of features for focused operations and policy recommendations in designated areas. The research is based on data collected from three regions of India about maternal health and maternal mortality. The results highlight the accuracy of healthcare operations, thereby facilitating data-informed decisions. Implementing the method outlined in this paper in any other region across the globe will significantly enhance the design and execution of targeted healthcare initiatives, enhancing public health outcomes and optimizing resource.
The task of finding associative entities in knowledge graph (KG) is to provide a ranking list of entities according to their association degrees. However, many entities are not only linked in KG but also associated in terms of user behaviors, which facilitates finding associative entities accurately. This manuscript incorporates KG with user-generated data to propose the Association Entity Graph Model (AEGM) to evaluate the association degrees. They first propose the joint weighting function to evaluate the entity associations and prove its submodularity theoretically as well as the greedy algorithm to select the candidates efficiently. They define the entity association information to score the entity association and give the hill climbing search based algorithm for AEGM construction. Following, they embed AEGM to calculate the association degrees and obtain the associative entities efficiently. Extensive experiments on three datasets show that the proposed method can achieve a better performance than some state-of-the-art competitors in accurately finding associative entities.
Artificial intelligence and machine learning have been transforming the health care industry in many areas such as disease diagnosis with medical imaging, surgical robots, and maximizing hospital efficiency. The Healthcare service market utilizing Artificial Intelligence is expected to reach 45.2 billion U. S. Dollars by 2026 from its current valuation, off $4.9 billion. Diabetic Retinopathy (DR) is a disease that results from complications of type one and Type two diabetes and affects patients' eyes. Diabetic retinopathy, if remains unaddressed, is one of the most serious complications of diabetes, resulting in permanent blindness. The disease has been affecting the lives of 347 million people worldwide. The paper aims to propose a residual network-based deep learning framework for the detection of diabetic retinopathy. The accuracy of our approach is 83% whereas the precision value for checking the absence of DR is 95%.
From the perspective of the inhibition of privacy computing theory, taking user satisfaction as risk perception, this paper discusses how the readability of financial APP privacy policies inhibits user satisfaction, thus affecting enterprise performance. Firstly, this paper measures the readability of privacy policies by constructing a professional vocabulary and an improved Fog Index formula. Secondly, this paper collects 615 APP privacy policies and corresponding enterprise data and empirically analyzes the impact of the readability of financial APP privacy policies on enterprise performance. This paper enriches research in the field of financial technology and empirically finds that the readability of financial technology APP privacy policies has a negative impact on enterprise performance, while user satisfaction has a positive impact on enterprise performance. In practice, this paper provides a reference for financial technology enterprises to improve the readability of APP privacy policies to improve user satisfaction and enterprise performance.
The acquisition and sharing of reviews have significant ramifications for the selection of crowdsourcing designs before mass production. This article studies the optimal decision of a brand enterprise regarding the acquisition/sharing of crowdsourcing design reviews in a supply chain. The authors consider an analytical model where the brand enterprise can privately acquire the manufacturer's review (MR) of crowdsourcing product designs and choose one of two information-sharing schemes—optional or mandatory sharing—to disclose MR to the key opinion leaders (KOLs), which help them to produce fans' reviews (FR). MR and FR integrate into the joint reviews (JR) that impact prospective consumers' purchase intention. The authors find that mandatory sharing significantly harms the brand enterprise's motivation to obtain MR, yet optional sharing is conducive to boosting JR on crowdsourcing designs. In addition, JR has a ceiling value, implying that excessively high FR and MR could not always enhance the effect of JR on crowdsourcing designs.
