7712 research outputs found
Sort by
Natural resource dependence and war nexus: new insights
No full textThis paper examines the channels through which resource dependence affects the probability of war. We consider all types of war within a worldwide panel dataset of all countries and territories, spanning the period 1960-2022. We systematically estimate the probability of war using a set of economic and geographic characteristics, including democracy, demography, legal origin, militarism, and sectoral composition. Using a panel probit model, we find that natural resources rents, military expenditures, and the presence of a common law tradition significantly increase the probability of war occurrence. However, we notice that this probability is decreasing in the level of economic development and democratization, and with the size of the services sector and trade openness. We also find that the effect of resource dependence varies by level of economic development, continent, landlocked status, legal origin, colonial history, and war type with civil wars being the salient type. Overall, we provide novel and updated evidence on an important dimension of the vexing question of the so-called ‘resource curse’
Cognitive AI and implicit pseudo-spline wavelets for enhanced seismic prediction
No full textUsing data from 1900 to 2024, this study developed an innovative artificial intelligence (AI)-powered framework for predicting earthquakes in Japan. By incorporating state-of-the-art cognitive computing techniques with expert seismic assessments, the proposed algorithm addresses some of the complex challenges in earthquake prediction. The model fuses AI systems with numerical methods such as the Finite Element Method (FEM) and pseudo-spline collocation techniques to simulate seismic wave propagation in a stratified spherical Earth. This study employed cognitive computing mechanisms to categorize and analyze seismic activities using a vector-based structure that compares past seismic events with predefined classifications. Moreover, the framework integrates expert knowledge of the stress distribution in the Earth\u27s crust to establish a comprehensive model for seismic forecasting. This AI-driven methodology provides deeper insight into seismic wave behavior and introduces a self-improving data-centric system that could support decision-making for reducing earthquake risk
Enhancing pro-environmental behavior in tourism: Integrating attitudinal factors and Norm Activation Theory
No full textThis research utilizes the Norm Activation Theory to explore tourists\u27 pro-environmental behavior. The study has two parts: the first qualitatively explores constructs predicting visitors\u27 pro-environmental behavior, while the second quantitatively measures pro-environmental behaviors and individual attitudes through a survey. Since augmented reality (AR) is transforming sustainable tourism by offering interactive educational experiences, during EXPO 2020 in Dubai, 1506 participants who engaged with AR were surveyed electronically from October 2021 to May 2022. Partial least squares structural equation modeling (PLS-SEM) was used to test hypotheses. The results suggest that tourists\u27 environmental values and sensitivity enhance their responsibility aspirations, with awareness of tourism\u27s negative environment and mitigation knowledge moderating the effects of responsibility aspirations on personal norms and pro-environmental behavior, respectively, while personal norms mediate the relationship between responsibility aspirations and pro-environmental behavior. These insights are valuable for tourism stakeholders, policymakers, and organizations aiming to promote environmental sustainability in the industry
The dual impact of audit partner busyness and boardroom gender diversity on audit quality in Australia
No full textPurpose: This study aims to examine the association between audit partner busyness and audit quality. Moreover, this research investigates whether boardroom gender diversity moderates the relationship between audit partner busyness and audit quality in Australia. Design/methodology/approach: The study sample consists of all public companies listed on the Australian Stock Exchange from 2005 to 2014. The data is obtained from SIRCA and the Morning Star databases. The study uses fixed effects and logistic regression techniques to test the relationship between audit partner busyness, boardroom gender diversity and audit quality. Findings: The collected empirical evidence shows that audit partner busyness is negatively associated with audit quality. In contrast, boardroom gender diversity moderates the relationship between audit partner busyness and audit quality. More specifically, the results suggest that board gender diversity mitigates the negative impact of audit partners’ busyness on the audit quality. The results are robust to endogeneity and alternative definitions of audit partner busyness, boardroom gender diversity and audit quality. Practical implications: The study’s findings will be of interest to policymakers, regulators and investors in the Australian market. The results show the importance of gender-diverse boards in companies’ audit functions, particularly in the presence of busy audit partners, and hence support the call for more women on corporate boards in Australia. Moreover, the results call for a cap or upper limit on the number of clients an audit partner can take on based on their capacity. Originality/value: The authors contribute to the growing literature on board gender diversity, audit partner busyness and audit quality. Although a plethora of prior literature suggests a negative association between audit partner busyness and audit quality, the results suggest that women in the boardroom positively moderate the relationship between audit partner busyness and audit quality
English‐Language Advertising in Europe
No full textResearch on advertising in Europe referring to the world Englishes framework has focused on the characteristics of English as it is used in different advertising genres in Europe, as well as on the attitudes that are held towards it, and the impact that its inclusion may have. It has included corpus studies that look at the incorporation of English into specific genres in different national or regional contexts, as well as experimental studies that have asked respondents to self‐report on their response to that English. Scholars have investigated television and radio commercials, print advertising and job advertisements, and studies have been carried out across the continent, from the Republic of Ireland in the West, to Russia in the East. This research identifies a number of shifts in the sociolinguistic landscape that are reflected in the ways in which English has been included in European advertising over the past two decades
Application of machine learning and deep learning techniques on reverse vaccinology – a systematic literature review
No full textReverse vaccinology (RV) is recognized as a productive method of vaccine discovery since it may be used to create vaccines for a variety of infectious pathogens. With the potential for machine learning (ML) algorithms to enable quick and precise predictions of vaccine candidates against new infections, RV is of particular relevance. Despite the fact that ML has been used successfully in the past, Deep learning (DL) model-based RV approaches have not been used widely. DL techniques are known to provide more complicated models and better performance for AI applications. This paper supports and reviews the roles of machine learning and Deep Learning in predicting potential vaccine candidates and discovery processes. Our study involved a systematic evaluation of selected publications, identified through a combination of prior knowledge and keyword searches across freely accessible databases. A meticulous screening process, considering contextual relevance, abstract quality, methodology, and full-text content, was employed. The literature review, conducted with a rigorous methodology, encompasses a thorough analysis of articles focusing on machine learning and deep learning techniques
Chapter 1 Cloud-to-things continuum: technical challenges and economical opportunities
No full textThis chapter introduces the concept of the cloud-to-thing continuum, a paradigm that integrates cloud computing and the Internet of Things (IoT) to enable a seamless and interconnected digital ecosystem. The chapter begins by discussing the emergence of IoT in cloud computing, highlighting the evolution from centralized cloud architectures to distributed edge computing. The technical challenges faced when implementing cloud-to-thing solutions are raised, including connectivity, data management, security, and interoperability. The impact of the cloud-to-thing continuum on digital transformation is emphasized via connected devices that are driving innovative business models and enhancing operational efficiencies. Furthermore, the chapter addresses the business challenges associated with the cloud-to-thing continuum, such as data ownership and governance, organizational silos, adoption barriers, and legal and regulatory hurdles. The chapter concludes by highlighting the potential of the cloud-to-thing continuum to revolutionize industries, enabling personalized services, and driving economic growth, while taking into account ethical considerations
Chapter 7 The Enterprise IoT
No full textDigital transformation in enterprises is driven by integrating AI and IoT, whereby companies maximize their return on investment and optimize the use of their operational resources. This chapter examines enterprise IoT, focusing on key aspects, such as scalability, connectivity, security, and data processing, in business contexts. The distinctions between consumer IoT and enterprise IoT are discussed to highlight the unique challenges businesses must consider when integrating IoT solutions into business processes. Case studies demonstrate how AI-enabled enterprise IoT models can optimize operations and personalize customer experiences. Additionally, these case studies illustrate new business models and revenue streams, which demonstrate the transformative potential of enterprise IoT across various industries
Unlocking the potential of EEG in Alzheimer\u27s disease research: Current status and pathways to precision detection
Full text linkAlzheimer\u27s disease (AD) affects millions of individuals worldwide and is considered a serious global health issue due to its gradual neuro-degenerative effects on cognitive abilities such as memory, thinking, and behavior. There is no cure for this disease but early detection along with a supportive care plan may aid in improving the quality of life for patients. Automated detection of AD is challenging because its symptoms vary in patients due to genetic, environmental, or other co-existing health conditions. In recent years, multiple researchers have proposed automated detection methods for AD using MRI and fMRI. These approaches are expensive, have poor temporal resolution, do not offer real-time insights, and have not proven to be very accurate. In contrast, only a limited number of studies have explored the potential of Electroencephalogram (EEG) signals for AD detection. In contrast, Electroencephalogram (EEG) signals present a cost-effective, non-invasive, and high-temporal-resolution alternative for AD detection. Despite their potential, the application of EEG signals in AD research remains under-explored. This study reviews publicly available EEG datasets, the variety of machine learning models developed for automated AD detection, and the performance metrics achieved by these methods. It provides a critical analysis of existing approaches, highlights challenges, and identifies key areas requiring further investigation. Key findings include a detailed evaluation of current methodologies, prevailing trends, and potential gaps in the field. What sets this work apart is its in-depth analysis of EEG signals for Alzheimer\u27s Disease detection, providing a stronger and more reliable foundation for understanding the potential role of EEG in this area
Assessment of cognitive domains in major depressive disorders using the Cambridge Neuropsychological Test Automated Battery (CANTAB): Systematic review and meta-analysis of cross-sectional and longitudinal studies
Full text linkCognitive difficulties are known to persist after remission of symptoms and to affect psychosocial functioning and quality of life. Cognitive function, measured with the Cambridge Neuro-psychological Test Automated Battery (CANTAB), is a reliable approach to measure cognitive function in major depression. This systematic review and meta-analysis appraise cross-sectional and longitudinal studies that used specific CANTAB tests to measure cognitive function in major depression and the effect of treatment (PROSPERO ID: CRD42022355903). 1212 studies were identified and 41 were included, 1793 patients and 1445 healthy controls. Deficits in executive functions were detected with the Stocking Of Cambridge (SOC) ‘number of problems solved with minimal number of moves’ and ‘subsequent thinking time’, Intra-Extra Dimensional Set Shift ‘number of trials to complete the test’, Spatial Working Memory ‘strategy score’ and ‘between errors score’, Spatial Span. Memory deficits were detected with Paired Associates Learning ‘number of total errors’, Pattern Recognition Memory (PRM) ‘% of correct answers’ and ‘response latency’, Spatial Recognition Memory ‘% of correct answers’, Delayed Matching To Sample (DMS) ‘% of total responses’. Impaired attention was detected by Rapid Visual Information Processing ‘response latency’ and probability to detect target’. Mental and motor responses increased when Reaction Time was measured. SOC ‘number of problems solved with minimal number of moves\u27, PRM ‘response latency’ and DMS ‘% of total responses\u27 improved after a course of treatment. A range of variables including year of publication, age, IQ, severity and duration of illness influenced cognitive changes. The presence of significant cognitive deficits requires novel targeted interventions