Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
8 results
Search Results
Item Restricted Human Action Recognition Based on Convolutional Neural Networks and Vision Transformers(University of Southampton, 2025-05) Alomar, Khaled Abdulaziz; Xiaohao, CaiThis thesis explores the impact of deep learning on human action recognition (HAR), addressing challenges in feature extraction and model optimization through three interconnected studies. The second chapter surveys data augmentation techniques in classification and segmentation, emphasizing their role in improving HAR by mitigating dataset limitations and class imbalance. The third chapter introduces TransNet, a transfer learning-based model, and its enhanced version, TransNet+, which utilizes autoencoders for improved feature extraction, demonstrating superior performance over existing models. The fourth chapter reviews CNNs, RNNs, and Vision Transformers, proposing a novel CNN-ViT hybrid model and comparing its effectiveness against state-of-the-art HAR methods, while also discussing future research directions.23 0Item Restricted ASSESSING THE IMPACT OF THE INTRODUCTION OF A RESEARCH-LED COMPUTER SCIENCE FRAMEWORK FOR PRIMARY EDUCATION IN THE KINGDOM OF SAUDI ARABIA(UNIVERSITY OF GLASGOW, 2024-02) Alharbi, Noha Abdulkhalig; Cutts, QuintinPrimary teachers around the world are being asked to teach computational thinking with little or no prior knowledge and limited support. This dissertation starts with analysing the challenges primary teachers face when teaching a subject new to them, identifying a lack of content and pedagogical content knowledge as a critical hurdle among many. Related situations in other subject areas are identified, where a new perspective on those subject areas has become central, and their approaches explored: the introduction of inquiry science and a more problem-based focus for mathematics both made use of a high-level framework that helped teachers to connect the top-level outcomes with low-level classroom materials provided to them. Computer science education is considered as a tool to develop computational thinking skills among learners, and there are worldwide efforts to implement it at the K-12 education level. However, being a relatively new subject, the teachers face similar challenges as the Mathematics and Science teachers mentioned above. Therefore, drawing on the mathematics/science experience, and on existing frameworks and research findings in CS, a Research-Led Computer Science Framework (RLCSF) that has three major components is presented. These components include the problem domain, computing domain, and problem-solving process, and computational thinking is presented as a modelling activity. The effectiveness of the framework was evaluated in the Kingdom of Saudi Arabia (KSA). To ensure that KSA is an appropriate evaluation context, data was collected from 114 teachers using the METRECC survey tool. This survey work provided detailed insight into the state of the intended and enacted curriculum at the K-12 level in KSA and the challenges the teachers face while teaching CS. The survey reports that the teachers in KSA have a limited understanding of computational thinking and problem-solving, and hence KSA is an appropriate context. For the evaluation, a Professional Development Program (PDP) was developed for teachers in which a teacher training guide was created, and a Professional Development Course (PDC) was conducted to educate the teachers about the way the RLCSF works and can solve the Content Knowledge (CK) and Pedagogical Content Knowledge (PCK) issues. As part of the PDP, other teacher training sessions were conducted during teaching. The researcher not only trained the teachers but also recorded their use of the framework and feedback using mixed-method approaches such as focus group studies, interviews and learners evaluation. To ensure the effectiveness of the results, another training session and interview were conducted with the teachers. The researcher involved one of the experienced group teachers, who is a non-CS background teacher from KSA, in the process of training and presentation. The objective is to investigate if the RLCSF can be transferred by a teacher who can teach the training processes and has recently experienced the framework. The second training guide is text-based, along with a video presentation explaining the examples given in the text guide. The teacher continued to assist teachers during their teaching. In the first semester, the CS teachers were able to use the materials more effectively than the non-CS teachers, but in the second semester, non-CS teachers were also at the same level. Teachers were able to start creating lessons using the material and framework. The results also reflect that the performance of learners from the experimental group was better than that of the control group learners. In the end, the non-CS teachers developed an understanding of why and how to develop computational thinking. The CS teachers had earlier focused only on teaching tools, but they developed an understanding of the importance of computational thinking. The teachers understand that modelling is a critical process in problem-solving. The results are promising and show that teachers are better able to understand different examples in the given curriculum and are able to deliver the contents more effectively.15 0Item Restricted Exploring Gamification by Evaluating Interfaces of Task Management Apps and Developing an Interface to Ensure Inclusiveness of Different Students’ Needs.(Newcastle University, 2024-09) Almakran, Raneem; Claisse, CarolineThe aim of this project is to evaluate and develop a prototype of a gamified task management app that enhances engagement for diverse students in academic contexts, such as students with ADHD symptoms. ADHD students face challenges with managing tasks, which affect their academic performance. Existing task management apps are designed for general users, but none specifically fulfill the needs of students with ADHD symptoms. Therefore, this paper evaluates two existing task management apps and presents a prototype designed to manage tasks in an enjoyable way by including gamification features such as competing with friends, collecting points, and taking care of a plant by achieving tasks. To understand user needs, semi-structured interviews were conducted with two students who identified with ADHD symptoms. Interviews were analysed using a qualitative approach. The results demonstrated the challenges faced in academia and how students tend not to use current task management apps for academic tasks. To test the usability of the gamified prototype, one participant was interviewed. Before the usability test, an ADHD persona and some scenarios were presented to the participant. The results were encouraging, suggesting that the gamified interface is engaging and enjoyable while managing tasks.50 0Item Restricted Visualising of cyber crime data by Communication Structured Acyclic Nets(Newcastle University, 2024-09-02) Alahmadi, Mohammed Saud; Koutny, MaciejCommunication Structured Acyclic Nets (CSA-nets) are a Petri net-based formalism used to represent the behaviour of Complex Evolving Systems (CES). CSA-nets, comprising sets of acyclic nets, are suitable tools for modelling and visualising the behaviour of event-based systems. Each subsystem is represented using a separate acyclic net, linked to others through a set of buffer places depicting their interactions. However, CSA-nets suffer from challenges especially in analysing and visualising CESs that have a large number of subsystems resulting from alternative and concurrent execution scenarios. Moreover, CSA-nets currently lack the capability to represent multiple or coloured tokens, thereby limiting their ability to represent several similar processes simultaneously. This thesis introduces extensions for CSA-nets to capture compactly the relationships between interacting systems’ components represented by sets of acyclic nets. Specifically, it introduces a way of folding buffer places to address the issue of a large number of buffer places. Then it introduces a new class of CSA-nets, called Parameterised Communication Structured Acyclic Nets (PCSA-nets), using multi-coloured tokens and allowing places to accept multiple tokens distinguished by parameters. The thesis also aims at improving the visualisation of csa-nets by rearranging their component acyclic nets to minimise the number of crossing arcs by taking inspiration from the main ideas behind three well-known sorting algorithms (bubble sort, insertion sort, and selection sort). Furthermore, this thesis presents a novel approach that combines TCP protocol anomaly detection with visual analysis through CSA-nets. The strategy provides a clear visualisation of cyber attack behaviours, leading a deeper understanding of Distributed Denial of Service (DDoS) patterns and their underlying causes. A new concept of Timed-Coloured Communication Structured Acyclic Nets (TCCSA-nets) is introduced, which allows elaboration of the system’s performance and emphasising the system’s operations in real-time. This approach allows for the classification of messages as abnormal if their duration exceeds a predetermined time limit.39 0Item Restricted EMPIRICAL EXPLORATION OF SOFTWARE TESTING(New Jersey Institute of Technology, 2024-04-18) Alblwi, Samia; Mili, Ali; Oria, VincentDespite several advances in software engineering research and development, the quality of software products remains a considerable challenge. For all its theoretical limitations, software testing remains the main method used in practice to control, enhance, and certify software quality. This doctoral work comprises several empirical studies aimed at analyzing and assessing common software testing approaches, methods, and assumptions. In particular, the concept of mutant subsumption is generalized by taking into account the possibility for a base program and its mutants to diverge for some inputs, demonstrating the impact of this generalization on how subsumption is defined. The problem of mutant set minimization is revisited and recast as an optimization problem by specifying under what condition the objective function is optimized. Empirical evidence shows that the mutation coverage of a test suite depends broadly on the mutant generation operators used with the same tool and varies even more broadly across tools. The effectiveness of a test suite is defined by its ability to reveal program failures, and the extent to which traditional syntactic coverage metrics correlate with this measure of effectiveness is considered.11 0Item Restricted Investigating Factors Influencing Blockchain Adoption in Saudi Healthcare Data Management(Florida Institute of Technology, 2024-05-15) Alkhalifah, Noura; Slhoub, KhaledBlockchain technology can potentially address security and privacy issues concerning the collection, storage, and sharing of healthcare data. However, its adoption within the healthcare sector is nascent in Saudi Arabia. This underutilization prompted our investigation into the determinants influencing blockchain adoption, intending to fully empower the Saudi healthcare sector to leverage blockchain capabilities. To achieve this, an extensive literature review was conducted to identify the pivotal factors encompassing technology, organization, and environment (TOE) that affect the successful implementation of blockchain technologies in managing healthcare data within the Saudi context. Utilizing the TOE framework, this study formulated three hypotheses concerning the adoption of blockchain technology. Subsequently, a quantitative analysis was undertaken through an online survey distributed among healthcare organizations in Saudi Arabia. We obtained responses from 129 valid ques- tionnaires and employed a partial least squares structural equation model (PLS-SEM) for analysis and hypothesis testing. The results show that technological and organizational factors significantly influence the adoption of blockchains, whereas environmental factors have no significance. This study contributes significantly to bridging a critical gap in the academic literature by clarifying the factors influencing blockchain adoption in healthcare data management in Saudi Arabia. Our findings serve as valuable guidelines for decision-makers contemplating the adoption of blockchain technology in healthcare data management, thus facilitating the effective navigation of associated challenges.22 0Item Restricted Deep Learning-Based Digital Human Modeling And Applications(Saudi Digital Library, 2023-12-14) Ali, Ayman; Wang, PuRecent advancements in the domain of deep learning models have engendered remarkable progress across numerous computer vision tasks. Notably, there has been a burgeoning interest in the field of recovering three-dimensional (3D) human models from monocular images in recent years. This heightened interest can be attributed to the extensive practical applications that necessitate the utilization of 3D human models, including but not limited to gaming, human-computer interaction, virtual systems, and digital twin. The focus of this dissertation is to conceptualize and develop a suite of deep learning-based models with the primary objective of enabling the expeditious and high-fidelity digitalization of human subjects. This endeavor further aims to facilitate a multitude of downstream applications that leverage digital 3D human models. The endeavor to estimate a three-dimensional (3D) human mesh from a monocular image necessitates the application of intricate deep-learning models for enhanced feature extraction, albeit at the expense of heightened computational requirements. As an alternative approach, researchers have explored the utilization of a skeleton-based modality, which represents a lightweight abstraction of human pose, aimed at mitigating the computational intensity. However, this approach entails the omission of significant visual cues, particularly shape information, which cannot be entirely derived from the 3D skeletal representation alone. To harness the advantages of both paradigms, a hybrid methodology that integrates the benefits of 3D human mesh and skeletal information offers a promising avenue. Over the past decade, substantial strides have been made in the estimation of two-dimensional (2D) joint coordinates derived from monocular images. Simultaneously, the application of Convolutional Neural Networks (CNNs) for the extraction of intricate visual features from images has demonstrated its prowess in feature extraction. This progress serves as a compelling impetus for our investigation into a hybrid architectural framework that combines CNNs with a lightweight graph transformer-based approach. This innovative architecture is designed to elevate the 2D joint pose to a comprehensive 3D representation and recover essential visual cues essential for the precise estimation of pose and shape parameters. While SOTA results in 3D Human Pose Estimation (HPE) are important, they do not guarantee the accuracy and plausibility required for biomechanical analysis. Our innovative two-stage deep learning model is designed to efficiently estimate 3D human poses and associated kinematic attributes from monocular videos, with a primary focus on mobile device deployment. The paramount significance of this contribution lies in its ability to provide not only accurate 3D pose estimations but also biomechanically plausible results. This plausibility is essential for achieving accurate biomechanical analyses, thereby advancing various applications, including motion tracking, gesture recognition, and ergonomic assessments. Our work significantly contributes to enhancing our understanding of human movement and its interaction with the environment, ultimately impacting a wide range of biomechanics-related studies and applications. In the realm of human movement analysis, one prominent downstream task is the recognition of human actions based on skeletal data, known as Skeleton-based Human Action Recognition (HAR). This domain has garnered substantial attention within the computer vision community, primarily due to its distinctive attributes, such as computational efficiency, the innate representational power of features, and robustness to variations in illumination. In this context, our research demonstrates that, by representing 3D pose sequences as RGB images, conventional Convolutional Neural Network (CNN) architectures, exemplified by ResNet-50, when complemented by as tute training strategies and diverse augmentation techniques, can attain State-of-the-Art (SOTA) accuracy levels, surpassing the widely adopted Graph Neural Network models. The domain of radar-based sensing, rooted in the transmission and reception of radio waves, offers a non-intrusive and versatile means to monitor human movements, gestures, and vital signs. However, despite its vast potential, the lack of comprehensive radar datasets has hindered the broader implementation of deep learning in radar-based human sensing. In response, the application of synthetic data in deep learning training emerges as a crucial advantage. Synthetic datasets provide an expansive and practically limitless resource, enabling models to adapt and generalize proficiently by exposing them to diverse scenarios, transcending the limitations of real-world data. As part of this research’s trajectory, a novel computational framework known as "virtual radar" is introduced, leveraging 3D pose-driven physics-informed principles. This paradigm allows for the generation of high-fidelity synthetic radar data by merging 3D human models and the principles of Physical Optics (PO) approximation for radar cross-section modeling. The introduction of virtual radar marks a groundbreaking path towards establishing foundational models focused on the nuanced understanding of human behavior through privacy-preserving radar-based methodologies.16 0Item Restricted Instructional Strategies for Teaching Computational Thinking in Secondary School Computer Science Introductory Courses(2023-05) Alghamdi, Khadijah Ali; Leftwich, Anne OttenbreitThere is a consistent call for teaching computational thinking (CT) in computer science (CS) courses and a massive body of CT research that emphasizes its importance in improving students’ computational skills. However, it is not clear how CT has been included and implemented in CS courses. This study explored how prevalent CT practices (abstraction, algorithms, pattern recognition, and decomposition) are in AP Computer Science Principles (AP CSP) courses and how they are implemented and taught. The study also examined teachers’ experiences and needs in teaching CT practices. I employed a mixed method multiple-case study design to explore the implementation of CT practices. Five AP CSP courses were analyzed (Code.org, Mobile CSP, BJC, Microsoft MakeCode, and UTeach). Also, five CS teachers who had at least one year of experience in teaching one of these courses were interviewed. For courses, frequencies and percentages of lessons that included CT practices were determined, and for quantitative data, thematic analysis procedures and constant comparative analysis procedures were employed. The results showed that abstraction was the most included, while decomposition was minimally included. In addition, this study pointed out different instructional strategies to teach abstraction and algorithms in an engaging environment. Teaching pattern recognition was mainly included in activities designed mostly to teach abstraction. However, decomposition was included mainly as a scaffolding strategy. The teachers reported improvement in their students’ understanding of abstraction. However, the teachers mentioned some challenges with teaching CT practices, such as student aversion to incorporate some CT practices in their coding, abstraction and algorithms in particular. Also, teachers mentioned that abstraction and decomposition are difficult for students. The teachers also reported that pattern recognition is not an easy CT practice and that students struggled to recognize patterns in their code. In addition, teachers agreed on the need for instructional strategies to teach pattern recognition and decomposition. It is recommended to explicitly teach pattern recognition and decomposition like what has been done with abstraction and algorithms. Teaching all four CT practices explicitly and as one entity is recommended. Also, showing the relationships between all CT practices as CT practices exist and how they can be used to solve problems is also recommended. More studies to focus on only one CT practice at a time or to examine CT practices in AP CSA curricula could be conducted in the future.46 0