Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
33 results
Search Results
Item Restricted SUCCESSION PLANNING IN THE PUBLIC SECTOR: A SYSTEMATIC REVIEW, MANAGERIAL IMPACT ANALYSIS, AND THE EMERGING ROLE OF ARTIFICIAL INTELLIGENCE IN HUMAN CAPITAL DEVELOPMENT(Saudi Digital Library, 2026) Alrawaf, Fahad Ibrahim Saad; Tantardini, MicheleSuccession planning is a critical strategic tool for public organizations to prepare and promote future leaders in response to leadership shortages. Despite its importance, research on its concepts, benefits, challenges, and influencing factors remains limited, as does understanding of its effectiveness and future potential, including the role of artificial intelligence (AI). This dissertation comprises three studies. The first study systematically reviews succession planning in public organizations, clarifying its core concepts, benefits, challenges, and success factors, while identifying gaps for future research. The second study quantitatively examines the impact of succession planning programs on work motivation, legacy motivation, job satisfaction, and turnover intentions, using Self-Determination Theory, Legacy Motivation Theory, and Social Exchange Theory. The analysis focuses on the National Program for Succession and Leadership Development, launched by the Ministry of Human Resources and Social Development in December 2024 and approved by the Saudi Council of Ministers in February 2025, which aligns with the Ambitious Nation pillar of Saudi Vision 2030. Results show positive effects on motivation and job satisfaction while reducing turnover intentions, introducing Legacy Motivation Theory as a new perspective. The third study employs a qualitative approach to explore HR professionals’ perceptions of AI in succession planning, guided by Human Capital Theory. Findings highlight AI’s potential to improve productivity, the need to develop digital competencies, and the importance of equity and transparency to avoid bias. Overall, the dissertation advances theoretical understanding and offers practical recommendations, emphasizing AI‑enabled succession planning, localized platforms to reduce algorithmic bias, ethical training, and strong governance frameworks to guide responsible AI adoption.19 0Item Restricted The Impact of AI Recommendation Features on Consumer Purchasing Decisions in E-Commerce Environments(Saudi Digital Library, 2026) Alrumi, Abdullah; Chipidza, WallaceArtificial intelligence recommendation features such as visual search, real-time offers, chatbot interactions, and generative recommendations have become central to e-commerce. A key gap remains because most studies focus on purchase intention instead of actual decisions and examine these AI features separately. This study addresses this gap by asking: What is the impact of AI recommendation features on consumer purchasing decisions in e-commerce environments? Using a sequential mixed-methods design in the Saudi Arabian e-commerce market, the quantitative phase surveyed 343 online shoppers analyzed with PLS-SEM, while the qualitative phase involved 15 interviews analyzed thematically with ATLAS.ti. The conceptual model was informed by the Technology Acceptance Model and Stimulus-Organism-Response framework. Results showed that trust was the only significant predictor of emotional reactions toward AI assistant tools use, while perceived intelligence, usefulness, and ease of use were not significant. Emotional reactions toward AI tools significantly influenced purchase decisions. The consideration set had a direct effect on purchase decisions but did not moderate the relationship. Qualitative findings revealed that AI does not simply assist at isolated points but actively shapes the entire buying journey from initial search to final decision. The significance appears in two areas. Theoretically, the study advances understanding of how AI-enabled systems shape consumer decision-making in digital platforms by focusing on purchase decisions rather than behavioral intentions and integrating quantitative and qualitative insights. Practically, the findings guide e-commerce businesses to design AI features that build consumer trust and support consumers throughout different stages of the buying journey.12 0Item Restricted Optimising care for acute ischaemic stroke: Early detection, treatment and outcomes(Saudi Digital Library, 2025) Alobaida, Muath Mubarak M; Lane, Deirdre; Lip, Gregory; Harrison, Stephanie; Rowe, FionaAbstract Background and Aim: Stroke management in pre-hospital and the early acute phase is crucial for timely and effective treatment to optimise care and outcomes. This thesis investigates the efficacy of machine learning (ML) models and traditional stroke scales for early detection of large vessel occlusion (LVO), evaluates novel visual impairment screening tools in emergency departments, and analyses outcomes of endovascular thrombectomy (EVT) in stroke patients with atrial fibrillation (AF). Methods: A systematic review and meta-analysis compared pre-hospital stroke scales and ML models for detecting LVO, where the ML models were based on clinical data (e.g., neurological examination findings and demographic characteristics). The effectiveness of Vision-Face-Arm-Speech-Time (V-FAST) checklist in detecting visual impairments was assessed against National Institutes of Health Stroke Scale (NIHSS) and orthoptist assessments in a hyperacute emergency setting. Outcomes of EVT in patients with AF were analysed, focusing on the impact of bridging thrombolysis (BT) and comparing sex and age differences in functional recovery and reperfusion success in a federated network and a nationwide cohort. Results: ML models showed higher discriminative performance than traditional stroke scales but faced challenges in real-world application due to variability and potential biases. The V-FAST checklist improved detection of several key visual impairments (visual field deficits, eye movement abnormalities, reading difficulties and visual extinction) in hyperacute settings. Its purpose is to serve as a screening tool to identify patients requiring comprehensive orthoptic assessment, although it was less effective in identifying complex eye movements disorders. AF status did not significantly impact haemorrhagic complications or mortality following EVT, and bridging thrombolysis (IV thrombolysis prior to EVT) offered a survival benefit in anticoagulated patients with AF. Females with AF had higher odds of good functional outcomes at 90 days compared to those without, and males with AF had higher successful reperfusion rates, especially in older groups. Conclusions: ML models can enhance early detection capabilities for LVO in pre-hospital settings, although their real-world application is limited by methodological and sample heterogeneity. The V-FAST checklist, evaluated within emergency department settings, shows improved detection of visual impairments in acute stroke care. AF status does not significantly impact EVT outcomes, supporting its safe use. Furthermore, the observed sex and age differences in EVT outcomes call for personalised stroke management approaches. Together, these findings present a cohesive research focus on improving acute stroke care across the entire patient journey, from early recognition and pre-hospital triage (via ML and stroke scales) to targeted symptom screening in emergency settings (V-FAST for visual deficits) and evaluation of outcomes in high-risk populations (AF patients undergoing EVT, including those on oral anticoagulants, and across demographic groups). This comprehensive approach supports the refinement of diagnostic and therapeutic protocols to enhance stroke care across diverse clinical environments and patient demographics.11 0Item Restricted Artificial intelligence for the detection and longitudinal monitoring of cardiovascular diseases(Saudi Digital Library, 2025) Alrumayh, Abdullah Ali; Peters, Nicholas; Bächtiger, PatrikHeart failure (HF) remains a major global health burden, contributing to substantial morbidity, mortality, and healthcare utilisation. Despite advances in cardiovascular care, early detection, accurate risk stratification, and long-term monitoring remain key challenges. Digital health technologies, particularly artificial intelligence (AI)-enabled diagnostics, offer potential solutions. AI-enhanced electrocardiography (AI-ECG) has shown promise in detecting left ventricular dysfunction (LVEF ≤40%), with potential applications beyond simple disease classification. This PhD systematically evaluates a single, pre-existing AI-ECG algorithm, focusing on its longitudinal prognostic value, risk assessment capacity, feasibility for self-administered remote monitoring (RM), and detection accuracy across diverse cardiovascular populations, with the overarching aim of bridging the gap between AI innovation and clinical implementation. Prospective multicenter studies assessed AI-ECG across clinical pathways. In newly diagnosed HF with reduced ejection fraction (HFrEF), AI-ECG probability scores correlated with LVEF trajectory and recovery, with each 10% increase in score associated with a shorter time to 10% LVEF improvement (adjusted hazard ratio [aHR] 1.71; p<0.001), supporting its role as a digital biomarker. AI-ECG-predicted LV dysfunction independently predicted major adverse cardiovascular events and all-cause mortality (aHR 1.93 and 1.56), even in patients with preserved LVEF. RM feasibility was demonstrated over 12 months, with 2,600 patient-collected ECGs and a real-time signal quality indicator enhancing engagement and data quality. Detection reliability was highest when ECG and LVEF were concurrent (AUROC 0.77), declining at 30 days (AUROC 0.62; p=0.0425), with superior performance in severe dysfunction versus borderline cases. External validation confirmed robust detection in newly diagnosed HFrEF (AUROC 0.86). Ongoing studies aim to expand AI-ECG’s applicability in chronic and acute cardiovascular care, tracking function, predicting complications, and optimizing treatments. In conclusion, this PhD advances AI-ECG as a tool for HF detection, risk stratification, and RM. Future work should prioritize large-scale validation, explainability, and integration strategies to ensure seamless adoption in clinical workflows.12 0Item Restricted Evaluating Machine Learning for Intrusion Detection in CAN Bus for in-Vehicle Security(Saudi Digital Library, 2025) Alfardus, Asma; Rawat, DandaThe past decade has seen a potential rise in the automobile industry accompanied by some serious challenges and threats. Increased demand for intelligent transportation system facilities has given a boom to the automotive industry. A safer and better experience is much sought from vehicles. It opens opportunities of including autonomous vehicles and Vehicle to Everything technologies in the automotive sector. Enabling vehicles to connect to various services exposes to compromise and misuse by the adversaries. There are numerous electronic devices in the modern vehicle which communicate with each other using multiple standard communication protocols. State-of-the-art vehicles are the assembly of complex mechanical devices with the sophisticated technology of electronic devices and connections to the external world. Controller Area Network (CAN) is one of the widely used protocols for in-vehicle communications. However, the lack of some fundamental security features such as encryption and authentication in CAN makes it vulnerable to security attacks. The backbone of connecting autonomous vehicles is CAN with limited bandwidth and exposure to unauthorized access. Various attacks compromise the confidentiality, integrity, and availability of vehicular data through intrusions which may endanger the physical safety of vehicles and passengers. These security shortcomings, therefore, lead to accidents and financial loss to the users of vehicles. To protect the in-vehicle electronic devices, researchers have proposed several security countermeasures. In this work, we discuss various security vulnerabilities and potential solutions to CAN’s. Further, a machine learning-based approach is also developed to devise an Intrusion Detection System for the CAN bus network. This study aims to explore the adaptability of the proposed intrusion detection system across diverse vehicular architectures and operational conditions. Furthermore, the findings contribute to advancing the state-ofthe-art in automotive cybersecurity, fostering safer and more resilient transportation ecosystems. Moreover, it investigates the scalability of the intrusion detection system to handle the increasing complexity and volume of data generated by modern vehicles.25 0Item Restricted Critical Success-Related Factors Influencing the Adoption and Use of Artificial Intelligence in Saudi Small and Medium-Sized Enterprises(Saudi Digital Library, 2025) alsulami, jehan fahim; Catherine, LouThe current decade presents significant opportunities for businesses to harness the transformative power of artificial intelligence (AI). Nonetheless, organisations of all sizes, including small and medium-sized enterprises (SMEs), continue to encounter challenges related to the critical success factors influencing AI adoption. Understanding the interplay between AI adoption, its utilisation, and these success factors remains pivotal to enhancing technology-enabled business operations. Therefore, this thesis investigates the critical success factors affecting AI adoption and use within Saudi SMEs. To construct a robust conceptual framework, three theoretical perspectives are employed based on a structured evaluation of relevant literature: the Technology-Organisation-Environment (TOE) framework, the Human-Organisation-and-Technology Fit (HOT-FIT) model, and Institutional Theory. The resulting framework addresses the technical, social, organisational, and environmental contexts critical to supporting SMEs in adapting to and utilising AI. It identifies eight key factors: skilled personnel, organisational readiness, data strategies, security concerns, system quality, government regulation, AI vendors, and trust in AI. A quantitative research methodology is employed to collect data from a sample of 300 SMEs across Saudi Arabia, utilising a simple random sampling technique within a cross-sectional survey design. Various statistical techniques are used to analyse and validate the proposed framework, including partial least squares structural equation modelling (PLS-SEM), which facilitates the testing and validation of the conceptual framework for AI adoption and use among Saudi SMEs. The findings indicate that critical success factors significantly impact AI adoption and use by Saudi SMEs. Notably, skilled personnel, government regulation, AI vendors, and trust in AI are identified as primary determinants of AI adoption. Additionally, skilled personnel, organisational readiness, government regulation, AI vendors, and trust in AI emerge as essential for the effective and sustainable use of AI. A statistically significant variation in AI adoption and usage is observed among Saudi SMEs of different sizes. The primary contribution of this study lies in extending existing information technology adoption literature to encompass the context of critical success factors for AI. Moreover, the findings offer a comprehensive perspective on the organisational dynamics influencing AI adoption and use by integrating human, organisational, technological, environmental, and social dimensions into a single framework. From a practical standpoint, the research provides valuable insights for technology consultants, policymakers, and regulatory authorities. Specifically, Saudi SMEs can leverage these findings to effectively enhance their capacity to adopt and sustain AI-driven innovations effectively.31 0Item Restricted Saudi Primary Teachers' Perceptions of Artificial Intelligence in Education: A Qualitative Investigation Through the TPACK-C Framework(Saudi Digital Library, 2025-06-28) Alshehri, Fahad; Marc, PruynSaudi Arabia's Vision 2030 identifies education as the principal driver of economic diversification and situates the Human Capability Development Program at the core of workforce preparation. Notwithstanding these ambitions, the 2018 PISA cycle showed that more than 50 percent of Saudi learners failed to attain minimum reading proficiency. The COVID-19 crisis subsequently demonstrated the system's capacity for rapid digital adaptation, characterised by agility and innovation. In this milieu, artificial intelligence has shifted from peripheral curiosity to policy imperative. No published study has yet employed the Technological Pedagogical Content Knowledge-Contextualised (TPACK-C) model enriched with Islamic constructs to investigate Saudi primary classrooms. This qualitative study explores five primary teachers' AI perceptions through TPACK-C, TAM, and SDT frameworks. Semi-structured Arabic interviews with teachers from diverse infrastructural contexts were analyzed using reflexive thematic analysis, revealing six key themes: instructional-efficiency catalyst, the infrastructure divide, variable student digital readiness, professional identity re-negotiation, fragmented professional learning, and the necessity of ethical-cultural safeguarding. Findings reveal a novel Threshold-Trajectory Model requiring sequential progression through digital-infrastructure, competency development, and ethical-cultural alignment phases. Results extend TPACK-C with Student Technological Knowledge (STK) and a culturally specific Ethical Knowledge domain, while providing actionable recommendations supporting Vision 2030 objectives.97 0Item Restricted ADVANCED LARGE LANGUAGE MODEL APPROACHES AND NATURAL LANGUAGE PROCESSING TECHNIQUES TO IMPROVE HATE SPEECH DETECTION USING AI(University of Central Florida, 2025) Almohaimeed, Saad; Boloni, LadislauThe proliferation of hate speech on social networks can create a significant negative social effect, making the development of AI-based classifiers that can identify and characterize different types of hateful speech in messages highly important for stakeholders. While this is a highly challenging problem, recent advances in language models promise to advance the state of the art such that even subtle and indirect forms of hate speech can be detected. In this dissertation we present a series of contributions that improve different aspects of hate speech classification. We developed THOS, a hate speech dataset consisting of 8.3k tweets. Compared to previous datasets, THOS contains fine-grained labels that identify not only whether a tweet is offensive or hateful, but also the target of the hate. Using this dataset, we studied the degree to which finer grained classification of tweets is feasible. In the follow-up work, we focus on the difficult problem of implicit hate speech, where hate is conveyed through subtle verbal constructs and allusions, without the use of explicitly offensive terms. We evaluate the efficacy of lexicon-based methods, transfer learning, and advanced LLMs such as GPT-4 on this problem. We found that the proposed techniques can boost the detection performance of implicit hate, although even advanced models often struggle with certain interpretations. In our third contribution, we introduce the Closest Positive Cluster (CPC) auxiliary loss, which improves the generalizability of classifiers across a wide range of datasets, resulting in enhanced performance for both explicit and implicit hate speech scenarios. Finally, given the scarcity of implicit hate speech datasets and the abundance of explicit hate datasets, we proposed an approach to generalize explicit hate datasets for the classification of implicit hate speech. Additionally, the proposed approach addresses noisy label correction issues commonly found in crowd-sourced datasets. Our method comprises three key components: influential sample identification, reannotation, and augmentation. We show that the approach improves generalization across datasets and enhances implicit hate classification.16 0Item Restricted Disinformation Classification Using Transformer based Machine Learning(Howard University, 2024) alshaqi, Mohammed Al; Rawat, Danda BThe proliferation of false information via social media has become an increasingly pressing problem. Digital means of communication and social media platforms facilitate the rapid spread of disinformation, which calls for the development of advanced techniques for identifying incorrect information. This dissertation endeavors to devise effective multimodal techniques for identifying fraudulent news, considering the noteworthy influence that deceptive stories have on society. The study proposes and evaluates multiple approaches, starting with a transformer-based model that uses word embeddings for accurate text classification. This model significantly outperforms baseline methods such as hybrid CNN and RNN, achieving higher accuracy. The dissertation also introduces a novel BERT-powered multimodal approach to fake news detection, combining textual data with extracted text from images to improve accuracy. By lever aging the strengths of the BERT-base-uncased model for text processing and integrating it with image text extraction via OCR, this approach calculates a confidence score indicating the likeli hood of news being real or fake. Rigorous training and evaluation show significant improvements in performance compared to state-of-the-art methods. Furthermore, the study explores the complexities of multimodal fake news detection, integrat ing text, images, and videos into a unified framework. By employing BERT for textual analysis and CNN for visual data, the multimodal approach demonstrates superior performance over traditional models in handling multiple media formats. Comprehensive evaluations using datasets such as ISOT and MediaEval 2016 confirm the robustness and adaptability of these methods in combating the spread of fake news. This dissertation contributes valuable insights to fake news detection, highlighting the effec tiveness of transformer-based models, emotion-aware classifiers, and multimodal frameworks. The findings provide robust solutions for detecting misinformation across diverse platforms and data types, offering a path forward for future research in this critical area.34 0Item Restricted A Systematic Review of User Consent, Transparency, and Secure Data Transmission and Storage(University of Technology Sydney (UTS), 2024-11-03) Alharbi, Sultanah; Hussain, Farookh KhadeerSmart home technology is revolutionizing residential environments by connecting devices to enhance comfort, safety, and energy efficiency. However, these advancements raise significant privacy concerns, particularly in data collection, transmission, and storage. This systematic review examines user consent, transparency, and secure data handling in smart homes, identifying challenges and innovative solutions such as blockchain and AI integration. The review highlights deficiencies in current consent mechanisms, the complexity of GDPR compliance, and practical barriers to implementation, offering insights for future research and practical privacy frameworks.32 0