Information Systems Frontiers最新文献

Data uncertainty is a challenging problem in machine learning. Distributionally robust optimization (DRO) can be used to model the data uncertainty. Based on DRO, a new support vector machines with double regularization terms and double margins can be derived. The proposed model can capture the data uncertainty in a probabilistic way and perform automatic feature selection for high dimensional data. We prove that the optimal solutions of this model change piecewise linearly with respect to the hyperparameters. Based on this property, we can derive the entire solution path by computing solutions only at the breakpoints. A solution path algorithm is proposed to efficiently identify the optimal solutions, thereby accelerating the hyperparameter tuning process. In computational efficiency experiments, the proposed solution path algorithm demonstrates superior performance compared to the CVXPY method and the Sequential Minimal Optimization (SMO) algorithm. Numerical experiments further confirm that the proposed model achieves robust performance even under noisy data conditions.

Government and public sector organizations focus on inclusiveness, transparency, security, and effectiveness. However, amid the ongoing deregulation and competition with private sectors, there are compulsions to leverage digital technologies for improving economic performance and customer satisfaction. In the above context, little is known about the role played by organizational capabilities in technology-led transformation of government. We employ SEM technique to validate our conceptual model that investigates the role of organizational capabilities on the degree of digitalization in the government organization and data for the same was collected from 196 government officials with relevant experience. The study found that effective digital strategy formulation increases employee digital capabilities and positively impacts digitalization in government. Vendor management capability is observed to mediate the relationship between digital strategy and digitalization in government while employee digital capability has an indirect impact through vendor management. Cultural and organizational barriers do not dampen the relationship between vendor management and digitalization in government. Our results integrate the dynamic capability framework with developments in digital transformation and technology sourcing literature.

Drawing on social capital theory, this study explores the antecedents of AI readiness in Small and Medium-sized Enterprises (SMEs) operating in resource-constrained environments, emphasising capabilities that mitigate cyber risks, and foster value construction in SMEs. Specifically, the study examines how structural, cognitive, and relational social capital fosters cyber resilience and contributes to proactive value construction, enhancing SMEs’ AI readiness and enabling them to construct and sustain value while safeguarding against potential cyber threats. The study adopts a Covariance-based Structural Equation Modelling (CB-SEM) approach to analyse 589 valid responses. A multi-wave data strategy with an interval cross-lagged design was implemented to reduce the risk of common method bias. The findings reveal that structural and relational capital significantly drive AI readiness, while cognitive social capital enhances cyber resilience, which is pivotal in constructing and protecting organisational value. Moreover, cyber resilience mediates the relationship between cognitive social capital and AI readiness, and enabling value construction amid cyber-related disruptions. SMEs with robust social capital networks are better equipped to leverage AI technologies for innovation and growth, construct new value streams, and defend against cyber risks, securing value in dynamic digital environments. This study contributes to the growing discourse on cybersecurity and digital transformation by offering insights into how SMEs can bolster digital innovation and construct sustainable value in the face of mounting cyber risks.

This study explores AI-enabled local public service provisioning, especially dilemmas of the mutual and meaningful development process. Theoretically, it builds on the current literature on AI implementation in the public sector and relates it to the theorisations of mutuality and meaningfulness. Empirically, it examines the experiences of public agents in a qualitative case study of chatbot development by the City of Oulu, Finland. The study concludes six factors that are constructed into three interconnected dilemma pairs to examine cross-cutting problematic decision-making scenarios and provide reconciliations through mutuality and meaningfulness.

A rich body of research examines the cybersecurity behavior of employees, with a particular focus on explaining the reasons why employees comply with (or violate) organizational cybersecurity policies. However, we posit that this emphasis on policy compliance is susceptible to several notable limitations that could lead to inaccurate research conclusions. In this research essay, we consider the limitations of using cybersecurity policy compliance as a dependent variable by presenting three assertions: (1) the link between policy compliance and organizational-level outcomes is ambiguous; (2) policies vary widely in terms of their clarity and completeness; and (3) employees have an inconsistent familiarity with their own organization’s cybersecurity policies. Taken together, we suggest that studying compliance with cybersecurity policies reveals only a partial picture of employee behavior. In response, we offer recommendations for future research.

Cybersecurity is now critically important in an increasingly digitized and connected world. In addition to required digital security, individuals and organisations pursue multiple other objectives under binding resource constraints. Understanding how they make decisions in the face of these trade-offs is important for both research and teaching purposes. Games can create effective and exciting learning environments and also provide an immersive and experiment-based research setting to understand decision-making. We present a novel tabletop board game which sets cybersecurity in a broader organisational context and emulates real life business decisions. It can be used as a powerful research tool to understand decision-making about cybersecurity in a resource-constrained and uncertain environment. It is also a useful interdisciplinary educational tool, integrating concepts from cybersecurity, business development, and innovation management in gameplay.

Given a data set and one single object known to be anomalous beforehand, the outlier explanation problem consists in explaining the abnormality of the input object with respect to the data set population. The approach pursued in this paper to solve the above task consists in finding an explanation, namely, a piece of information encoding the characteristics that locate the anomalous data object far from the normal data. Our explanation consists of two components, the choice, encoding the set of features in which the anomalous object deviates from the rest of the population, and the mask, encoding the associated amount of deviation with respect to the normality. The goal here is not to explain the decisional process of a model but, rather, to provide an explanation justifying the output of the decisional process by only inspecting the data set on which the decision has been made. We tackle this problem by introducing an innovative deep learning architecture, called MMOAM, based on the adversarial learning paradigm. We assess the effectiveness of our technique over both synthetic and real data sets and compare it against state of the art outlier explanation methods reporting better performances in different scenarios.

This paper examines the evolving dynamics of human-AI interaction, emphasizing the ethical challenges and responsibility gaps that emerge as AI technologies become more autonomous and integrated into society and business. We analyze, utilizing a systematic literature review, how various ethical views influence our understanding of morality and responsibility in human-AI collaborations. Deontological ethics emerge as a dominant theme, with much of the literature centered on ethical principles shaped by powerful nations. The study highlights the need to integrate diverse ethical perspectives into AI research to address contradictions in ethical frameworks across various cultural contexts. While respecting cultural differences, achieving a common ground among these frameworks requires increased dialogue among AI researchers and practitioners. Our findings further underscore the importance of future research in developing a more cohesive understanding of how AI transformation challenges previous assumptions about AI’s role in moral agency and responsibility.

The global commitment to advancing financial inclusion (FI) relies on technology to connect underserved communities with the formal financial sector. Existing traditional technologies have made some progress, but they often fail to adapt to the unique needs of these populations. Although artificial intelligence (AI) offers new possibilities to meet these limitations, its rapid advancement has outpaced the development of integrative studies, leaving its potential impact on financial access by the underserved largely unexplored. Existing research provides fragmented insights into how different technological interventions impact diverse groups. We develop a segment-outcome-focused analysis to structure a scoping review of 95 information systems studies to assess current technological advances for FI and explore how AI can address their limitations, including limited digital literacy, uneven and high cost of infrastructure, and service personalization. We then engender future research directions and conclude with theoretical contributions and practical implications, emphasizing the potential of AI solutions to advance FI.

The main objective of this study is to explore the emerging patterns in the Information Systems (IS) meta-analysis landscape by reviewing a sample of 162 studies published in IS journals and conferences covering 37 years of meta-analysis research. The findings underline the diversity of topics, bodies of knowledge, as well as theories and constructs on various levels of analysis that were examined in recent decades. The key findings and recommendations help IS scholars to conceptualize their meta-analysis study design and redirect their attention to under-researched areas and methodological issues that need improvement, thus to learn from the past for advancing the future.