Publication Search

The rapid growth of big data has significantly increased the demand for efficient and scalable data processing methods, particularly within cloud computing environments. This study aims to evaluate the effectiveness of distributed computing frameworks, specifically Apache Hadoop and Apache Spark, in optimizing big data processing. A qualitative approach using a Systematic Literature Review (SLR) method is employed to analyze existing studies related to distributed systems, cloud computing architectures, and performance optimization techniques. The analysis focuses on key performance indicators, including processing speed, resource utilization, and scalability, as well as the suitability of each framework for different data processing scenarios. The findings indicate that Apache Hadoop is highly effective for batch processing and storage-intensive tasks due to its disk-based architecture, while Apache Spark demonstrates superior performance in real-time and iterative processing through its in-memory computing capabilities. Additionally, system configuration factors such as cluster size, memory allocation, and network bandwidth are identified as critical elements influencing overall performance. The study also highlights emerging trends, including the adoption of hybrid cloud environments, the integration of artificial intelligence and machine learning, and the utilization of edge computing to enhance real-time data processing. In conclusion, distributed computing frameworks play a vital role in improving the efficiency and scalability of big data processing in cloud environments. The selection of an appropriate framework, combined with optimized system configuration, can significantly enhance operational performance and support data-driven decision-making.

Cloud-based resource allocation and VM/container orchestration play a crucial role in ensuring performance, scalability, and energy efficiency in modern distributed computing environments. This study investigates the effectiveness of centralized and decentralized scheduling models combined with heuristic and optimization-based allocation strategies in container-based cloud infrastructures. A quantitative experimental approach was employed to evaluate system performance under varying workload intensities. Key evaluation metrics included response time, throughput, resource utilization, SLA violation rate, and energy consumption. The experimental results indicate that centralized scheduling mechanisms experience scalability limitations and increased latency under high workload conditions. Although optimization-based allocation improves performance within centralized architectures, coordination bottlenecks remain significant. In contrast, decentralized scheduling models demonstrate superior adaptability, reduced response time, and improved throughput due to distributed decision-making and reduced control overhead. The integration of intelligent optimization techniques further enhances resource utilization and energy efficiency, achieving the lowest SLA violation rates and highest system stability. Overall, the findings confirm that combining decentralized scheduling with optimization-driven resource allocation provides a more scalable and sustainable orchestration strategy for modern cloud environments. This approach is particularly suitable for dynamic, large-scale, and latency-sensitive applications in container-based and edge-integrated cloud systems.

Cyber Physical Systems (CPS) are vital for managing and controlling critical infrastructures, such as industrial control systems, power grids, and transportation networks. These systems integrate digital and physical components, offering numerous benefits for industrial automation. However, the increasing interconnectivity of these systems has introduced new security vulnerabilities, particularly in anomaly detection and system reliability. This research aims to address these challenges by proposing an edge based anomaly detection framework that leverages lightweight deep learning models, specifically designed to operate efficiently on resource constrained edge devices. Literature Review: Previous studies have shown the effectiveness of anomaly detection in CPS, with traditional methods struggling to keep up with the complexity and scale of modern industrial environments. Machine learning and deep learning approaches, particularly hybrid models combining rule based systems and AI, have emerged as effective solutions for real time anomaly detection. Techniques such as model compression, quantization, and pruning are essential for adapting these models to resource limited edge devices while maintaining high detection accuracy and low latency. Materials and Method: The proposed framework integrates deep learning models such as Convolutional Neural Networks (CNNs) and Long Short Term Memory (LSTM) networks, optimized for edge computing environments. The datasets used for training and testing include industrial network traffic and sensor anomaly datasets. Model optimization techniques like pruning and quantization were applied to reduce computational overhead and energy consumption on edge devices. Results and Discussion: The framework demonstrated high detection accuracy (AUC of 0.9720) with ultra low latency (0.0019 seconds training time), making it highly suitable for real time anomaly detection in CPS. Resource efficiency was achieved by optimizing the models for edge devices, reducing energy consumption while maintaining performance. The framework also significantly improved security by identifying anomalies early, preventing potential threats to critical infrastructures. Future directions include exploring federated learning to enhance privacy and data sharing across distributed devices.

Collaborative filtering systems rely heavily on matrix factorization techniques, which often face scalability issues when handling large datasets. This paper presents an efficient parallel algorithm that leverages distributed computing to perform largescale matrix factorization. Experimental results show that our algorithm significantly reduces computation time while maintaining high accuracy. The approach has practical implications for recommendation systems, particularly in ecommerce and social media platforms.

The rapid growth of cross-platform applications has significantly increased the volume and diversity of sensitive user data processed across heterogeneous and distributed environments. Personally identifiable information, device identifiers, behavioral data, and financial information are routinely collected to support personalization, analytics, and service optimization. While these practices enhance application functionality and user experience, they also introduce substantial privacy risks, including unauthorized data access, device fingerprint–based re-identification, cross-user data leakage, and large-scale data breaches. These risks are further amplified by distributed processing architectures and extensive third-party library integrations commonly used in modern cross-platform systems. This study aims to systematically analyze privacy issues in cross-platform applications by examining the types of sensitive data involved, identifying dominant privacy threats, and reviewing state-of-the-art privacy-preserving mitigation strategies. A systematic literature-based methodology was employed, focusing on recent Scopus-indexed journal articles, conference papers, and book chapters. The analysis synthesizes findings using thematic categorization and a conceptual research framework that maps sensitive data sources to privacy threats and corresponding mitigation mechanisms. The results indicate that privacy risks in cross-platform applications originate not only from external attacks but also from internal architectural weaknesses, such as flawed authorization logic and excessive data sharing across system components. Privacy-preserving techniques including differential privacy, federated learning, blockchain-based data governance, secure multi-party computation, and fine-grained access control mechanisms are shown to provide stronger privacy guarantees compared to conventional centralized approaches. However, these techniques also present trade-offs related to system complexity and performance. Overall, the study highlights the importance of adopting a multi-layered, privacy-by-design approach to ensure sustainable, trustworthy, and regulation-compliant cross-platform application development.

The rapid growth of Internet of Things (IoT) and edge computing technologies has introduced new security challenges due to the distributed, heterogeneous, and dynamic nature of these environments. Conventional static security mechanisms, such as rulebased authentication and fixed trust models, are often inadequate for addressing evolving threats and abnormal behaviors in largescale IoT systems. To overcome these limitations, this study proposes a machine learningbased trust evaluation framework for enhancing security in distributed IoT environments. The proposed approach dynamically assesses the trustworthiness of IoT nodes by analyzing behavioral and interactionbased features collected at the edge layer. Machine learning models are trained to classify nodes into trusted and malicious categories and continuously update trust values in response to changing network conditions. Based on the predicted trust levels, adaptive security decisions are enforced to allow or restrict node participation in data sharing and computation processes. A quantitative experimental evaluation is conducted using simulated distributed IoT scenarios that include both normal and malicious behaviors. The performance of the proposed framework is evaluated using standard metrics such as accuracy, precision, recall, F1score, and detection effectiveness, and is compared against conventional static trust and rulebased security mechanisms. The results demonstrate that the proposed machine learningbased trust evaluation approach achieves significantly higher detection accuracy and robustness while maintaining low computational overhead. Overall, the findings confirm that integrating machine learning into trust management provides an effective and scalable solution for securing distributed IoT systems under dynamic and adversarial conditions.

By implementing the case tracking information system at the Special Class 1A Sidoarjo District Court, institutions can meet their demands for a fast judicial process as the world of justice continues to develop. Likewise with the case tracking information system implemented at the Special Class 1A Sidoarjo District Court. This information system is a website-based information system that can be accessed locally by internal courts. The aim of the research is to determine the level of satisfaction of SIPP users using the end-user satisfaction computing model approach which has been modified by Aggelidis & Chatzoglou and what factors of end-user computing satisfaction influence user satisfaction. The method in this research was collected using a questionnaire distributed to 70 respondents. The results of the research were that the level of satisfaction of SIPP users was 6% very satisfied, 81% felt satisfied, 12% felt neutral, 1% felt dissatisfied, 0% felt very dissatisfied. The results of the analysis obtained an R-square value of 79.1%, which states that information quality, system quality and support in sourcing have a positive and significant effect on overall satisfaction. Then system quality has a positive and significant effect on information quality. Then support in sourcing has a positive and significant effect on system quality