Information Systems Frontiers最新文献

As artificial intelligence (AI) and machine learning (ML) advance, concerns about their sustainability impact grow. The emerging field "Sustainability of AI" addresses this issue, with papers exploring distinct aspects of ML’s sustainability. However, it lacks a comprehensive approach that considers all ML development phases, treats sustainability holistically, and incorporates practitioner feedback. In response, we developed the sustainable ML design pattern matrix (SML-DPM) consisting of 35 design patterns grounded in justificatory knowledge from research, refined with naturalistic insights from expert interviews and validated in three real-world case studies using a web-based instantiation. The design patterns are structured along a four-phased ML development process, the sustainability dimensions of environmental, social, and governance (ESG), and allocated to five ML stakeholder groups. It represents the first artifact to enhance each ML development phase along each ESG dimension. The SML-DPM fuels advancement by aggregating distinct research, laying the groundwork for future investigations, and providing a roadmap for sustainable ML development.

Osteoarthritis (OA) is a common joint disease affecting people worldwide, notably impacting quality of life due to joint pain and functional limitations. This study explores the potential of radiomics — quantitative image analysis combined with machine learning — to enhance knee OA diagnosis. Using a multimodal dataset of MRI and CT scans from 138 knees, radiomic features were extracted from cartilage segments. Machine learning algorithms were employed to classify degenerated and healthy knees based on radiomic features. Feature selection, guided by correlation and importance analyses, revealed texture and shape-related features as key predictors. Robustness analysis, assessing feature stability across segmentation variations, further refined feature selection. Results demonstrate high accuracy in knee OA classification using radiomics, showcasing its potential for early disease detection and personalized treatment approaches. This work contributes to advancing OA assessment and is part of the European SINPAIN project aimed at developing new OA therapies.

Digital Twin (DT) technology monitors, simulates, optimizes, models, and predicts the behavior of physical entities. Healthcare is a significant domain where a DT can be functional for multiple purposes. However, these diverse uses of DTs need a clear understanding of both general and specific aspects that can affect their adoption and integration. This paper is a meta-review that leads to the development of a conceptual framework designed to support the high-level evaluation of DTs in healthcare. Using the PRISMA methodology, the meta-review synthesizes insights from 20 selected reviews out of 1,075 studies. Based on this comprehensive analysis, we extract the functional, technological, and operational aspects that characterize DTs in healthcare. Additionally, we examine the structural (e.g., hierarchical) relationships among these aspects to address the various complexity scales in digital health. The resulting framework can promote the effective design and implementation of DTs, offering a structured approach for their assessment.

Blockchain has become a promising technology with huge benefits; nevertheless, its adoption intention has been scarce across different organizations, especially in government-to-citizens service (e.g., blockchain-based e-voting). Therefore, we aim to investigate how blockchain can affect citizens' adoption intention to use blockchain-based e-voting service. We study blockchain adoption intention by employing UTAUT2 as a theoretical base and the Replacement-Amplification-Transformation (R.A.T) technology integration model to study digital literacy as a mediating mechanism in our study model due to its significance in the contemporary business world. On the method side, we obtained 315 valid responses that we utilized to conduct a PLS-SEM-based analysis. Our findings state that digital literacy positively mediates the relationship between five determinants of UTAUT2 (e.g., facilitating conditions, social influence, hedonic motivation, habit, and price value) and citizens' intention to adopt blockchain e-voting service for casting their votes in elections. This study is among the first to examine the mediating mechanism of digital literacy between UTAUT2 factors and citizens' intention to adopt blockchain e-voting service. It is also worthwhile to quote that our study is a pioneer in extending the UTAUT2 model in the context of blockchain e-voting service. Lastly, we communicate the study's theoretical and practical implications to enrich both knowledge and industry.

Accurate estimates of medication rankings and volumes can benefit patients, physicians, online health communities, pharmaceutical companies, and the healthcare industry at large. This study analyzes patient-generated content in online health communities to discover important medication transition and combination patterns for better ranking and volume predictions. The current research takes a data-driven analytics approach to identify medication information from patient posts and classify different types of medication relations. The identified relation patterns then are represented in a medication relation network with an adjusted transition model for ranking and volume estimates. Evaluation results of real-world patient posts show the proposed method is more effective for predicting medication rankings than existing network-based measures. Moreover, a regression-based model, informed by the proposed method’s network-based outcomes, attains promising accuracy in estimating medication volumes, as revealed by the relatively low mean squared errors. Overall, the proposed method is capable of identifying important features for increased predictive power, as manifested by ({text{R}}^{2}) and adjusted ({text{R}}^{2}) values. It has the potential for better medication ranking and volume predictions, and offers insights for decision making by different stakeholders. This method is generalizable and can be applied in important prediction tasks in healthcare and other domains.

As the frequency of jamming attacks on wireless networks has increased, conventional local jamming detection methods cannot counter advanced jamming attacks. To maximize the jammer detection performance of machine learning (ML)-based detection methods, a global model that reflects the local detection results of each local node is necessary. This study proposes an ML-based cooperative clustering (MLCC) technique aimed at effectively detecting and countering jamming in beyond-5G networks that utilize smart repeaters. The MLCC algorithm optimizes the detection rate by creating and updating a global ML model based on the jammer detection results determined by each local node. The network performance is optimized through load balancing among the smart repeaters and access points, and the best path is selected to avoid jammers. The experimental results demonstrate that the MLCC improves the detection rate and throughput by up to 5.21% and 26.35%, respectively, while reducing the energy consumption and latency by up to 76.68% and 7.14%, respectively.

Actions that people aim to do are considered one of the main drivers behind purchase decisions and uncovering people’s needs in a human-centered manner. Such actions are often expressed by buyers in product reviews. However, most existing recommender system approaches still lack incorporating buyer-product action knowledge in the recommendation process. This limitation increases the gap between buyers’ needs and the recommended products. This research proposes a knowledge graph-based framework to represent buyers’ action knowledge from product reviews and integrate it in recommender systems to provide more human-centered and explainable recommendations. The framework is validated through a set of prototypes, which demonstrate the feasibility of buyers expressing their needs in the form of actions and recommending products accordingly. An initial evaluation revealed a promising 75% System Usability Scale score, with interview-based feedback that shed light on the capabilities of the proposed approach in supporting buyers in their online product selection experience.

Dealing with privacy threats and adopting appropriate strategies to manage personal information have become crucial challenges for internet users. While adaptive problem-focused coping (APFC) has been extensively discussed in the literature on information privacy, little is known about maladaptive emotion-focused coping (MEFC). This paper proposes that individuals employ privacy protection motivation (a form of APFC) and privacy cynicism (a form of MEFC), according to their threat and coping appraisals. These two coping strategies will then influence their behaviour regarding the disclosure of personal information on the internet. Offering an empirical analysis of 346 samples of survey data from China, this paper reveals that privacy cynicism, which is mainly affected by deep concerns about privacy and high self-efficacy but low response efficacy, and inconsistency between users’ motivations for protecting their privacy and their actual disclosure behavior are the reasons for the privacy paradox. This study provides crucial theoretical support and practical guidance for the privacy management of internet users’ information.

Imbalanced data sets are a growing problem in data mining and business analytics. However, the ability of machine learning algorithms to predict the minority class deteriorates in the presence of class imbalance. Although there have been many approaches that have been studied in literature to tackle the imbalance problem, most of these approaches have been met with limited success. In this study, we propose three methods based on a wrapper approach that combine the use of under-sampling with ensemble learning to improve the performance of standard data mining algorithms. We test our ensemble methods on 10 data sets collected from the UCI repository with an imbalance ratio of at least 70%. We compare their performance with two other traditional techniques for dealing with the imbalance problem and show significant improvement in the recall, AUROC, and the average of precision and recall.

The recent advancements in the field of Artificial Intelligence (AI) have sparked a renewed interest in how organizations can potentially leverage and gain value from these technologies. Despite the considerable hype around AI, recent reports indicate that a very small number of organizations have managed to successfully implement these technologies in their operations. While many early studies and consultancy-based reports point to factors that enable adoption, there is a growing understanding that adoption of AI is rather more of a process of maturity. Building on this more nuanced approach of adoption, this study focuses on the diffusion of AI through a maturity lens. To explore this process, we conducted a two-phased qualitative case study to explore how organizations diffuse AI in their operations. During the first phase, we conducted interviews with AI experts to gain insight into the process of diffusion as well as some of the key challenges faced by organizations. During the second phase, we collected data from three organizations that were at different stages of AI diffusion. Based on the synthesis of the results and a cross-case analysis, we developed a capability maturity model for AI diffusion (AICMM), which was then validated and tested. The results highlight that AI diffusion introduces some common challenges along the path of diffusion as well as some ways to mitigate them. From a research perspective, our results show that there are some core tasks associated with early AI diffusion that gradually evolve as the maturity of projects grows. For professionals, we present tools for identifying the current state of maturity and providing some practical guidelines on how to further implement AI technologies in their operations to generate business value.