In the era of big data, clustering and data mining have become essential tools for uncovering patterns and insights from vast datasets. However, these processes often involve the use of sensitive data, raising significant concerns about privacy, security, and trustworthiness. This paper proposes N2P-CM, a novel privacy-preserving framework designed to protect sensitive information during the entire clustering and mining lifecycle. Unlike existing methods that focus on partial aspects of security or apply generic encryption techniques, N2P-CM integrates five innovative and synergistic modules: Sensitive Feature Obfuscation, Adaptive Trust Weight Aggregation, Compressed Secure Semantic Embedding, Differential Traceable Execution Engine, and Blockchain Auditable Ledger. Each module contributes a distinct layer of privacy and accountability, ranging from feature-level data transformation and federated trust scoring to secure semantic encoding and traceable execution logging with blockchain support. We provide formal definitions and algorithms for each module and demonstrate their integration in a unified architecture. Extensive simulations using real-world datasets validate the efficacy of N2P-CM, showing that it achieves strong privacy guarantees with minimal degradation in clustering accuracy. This research contributes a comprehensive and modular solution to the growing need for privacy-preserving analytics in sensitive domains such as healthcare, finance, and smart cities.

The purpose of this study was to examine the effects of key organizational readiness dimensions and extrinsic motivation on the teaching staff’s intention toward the continued voluntary use of e-learning innovations post COVID-19 pandemic. These effects have not received considerable focus in the extant e-learning literature. To mitigate this lack, an integrated model encompassing dimensions from several organizational readiness frameworks and a motivational theory was developed. The model postulated these dimensions as direct determinants of the e-learning innovations continuance intention. A structured questionnaire-based survey was conducted to empirically assess the developed model. The intended population for this survey was composed of teaching staff at a Saudi higher education institution characterized by a wide adoption of e-learning innovations during the pandemic. The 233 valid responses obtained from this population were analyzed using the structural equation modeling method. The results indicated that only two organizational readiness dimensions (i.e., teaching staff readiness and administrative support) and extrinsic motivation were significant positive drivers of the continuance intention to use e-learning innovations. According to these findings, the study emphasizes that the key e-learning stakeholders should develop effective policies and procedures that reinforce the roles of the examined dimensions in promoting such continuance intention, which represents a crucial indicator for the successful implementation of the adopted innovation.

Smart home technologies have revolutionized modern living by enhancing convenience, efficiency, and security. In contrast, many interconnected devices introduce significant security and privacy challenges. This comprehensive review investigates the integration of blockchain technology as a robust solution for secure access control in smart home environments. The decentralized and tamper-resistant nature of blockchain technology effectively solves important problems, including device authentication, data integrity, and access management, through the use of cryptography and distributed ledgers. The study synthesizes findings from 52 research papers, categorizing them into three thematic areas: blockchain in access control systems, its applications in IoT, and specific implementations for smart homes. It highlights the transformative potential of blockchain in mitigating vulnerabilities inherent in centralized systems, fostering trust, and enhancing security frameworks. Despite its promising applications, challenges such as scalability, interoperability, and energy consumption persist, warranting further research. This paper stresses the necessity of collaboration to tackle these limitations and enhance blockchain-based access control solutions for smart homes, setting the stage for more secure and user-focused smart environments.

Graphs have long been used to model relationships between entities. For some applications, a single graph is sufficient; for other problems, a collection of graphs may be more appropriate to represent the underlying data. Many contemporary problem domains, for which graphs are an ideal data model, contain an enormous amount of data (e.g., social networks). Hence, researchers frequently employ parallelized or distributed processing. The graph data must first be partitioned and assigned to the multiple processors in a way that the workload is balanced and inter-processor communication is minimized. The latter problem may be complicated by the existence of edges between vertices in a graph that have been assigned to different processors. Herein we introduce a strategy that combines vocabulary-based summarization of graphs (VoG) and detection of hotspots (i.e., vertices of high degree) to determine how a single undirected graph should be partitioned to optimize multi-processor load balancing and minimize the number of edges that exist between the partitioned subgraphs. We benchmark our method against another well-known partitioning algorithm (METIS) to demonstrate the benefits of our approach.

This paper aims to analyze the capability of advanced AI as a predictor of audit quality with particular reference to auditors in Kuwait. The research focuses on understanding the role of advanced AI technologies in the improvement of most audit activities around risk, fraud, and compliance. In order to classify the Kuwaiti auditors into different segments on the basis of their internet usage, both the quantitative data collected through a questionnaire survey is used with additional data collected from structured interviews with them. The results are expected to offer a rich and detailed account of the pragmatic opportunities and difficulties of applying AI in audits while highlighting its potential of reshaping conventional approaches. This study contributes relevant knowledge regarding the audit quality and governance garnered from linking theory and practice, providing the feasible recommendations for auditors and policymakers in the member countries of the GCC.

The use of technology in the financial industry has experienced sustained growth in recent years. However, in many emerging economies, a significant proportion of the population still does not utilize digital solutions for financial transactions. Promoting financial inclusion through digital environments is essential for driving social and economic development. This study aims to develop machine learning models to predict the adoption of digital payments in Latin America and the Caribbean using statistical data from the World Bank's Global Findex Database for 2021. The performance of the Random Forest, LightGBM, XGBoost, and CatBoost algorithms was compared, with the optimal hyperparameter combination identified through Bayesian optimization. The results show that LightGBM achieved the highest performance in predicting digital payments, with an F1-score of 90.25% and a more stable balance between precision and recall compared to the other models. These findings highlight the value of machine learning models in the financial sector, as they enable a more accurate identification of users adopting digital solutions, facilitating the design of strategies to strengthen financial inclusion in the region.

The main objective of spatiotemporal analysis is to offer precise predictions of outcomes. The objective of this study is to assess the accuracy of the Bayesian Latent Gaussian Model in predicting outcomes by utilizing both time-varying and fixed spatial weight matrices. The results of the Monte Carlo simulation suggest that when there is moderate spatial autocorrelation (between 0.3 and 0.7), it is strongly advised to use a time-varying spatial weight matrix. This approach yields the most precise predictions and minimizes any distortion in parameter estimates. Furthermore, we provide an illustrative case study where we simulate the effects of exposure to multiple pollutants on tuberculosis. The analysis revealed that particulate matter 10 (PM10), nitrogen oxides (NO2), sulfur dioxide (SO2), carbon monoxide (CO), and ozone (O3), have a positive influence on the risk of TB, with spatial effects that change over time. The model demonstrates that a rise of 1 mg/m³ in the levels of PM10, NO2, SO2, CO, and O3 is linked to corresponding increases in TB cases by 2.1%, 21.17%, 13.20%, 6.72%, and 6.59%, respectively. NO2 and SO2 have the most significant influence on the risk of tuberculosis (TB). These findings enhance our comprehension of the spatial correlation of TB over time and promote further investigation to determine the most efficacious strategies for mitigating the dissemination of TB.

Collecting poverty data through the National Socio-Economic Survey (SUSENAS) demands significant time, costs, and human resources. To enable more efficient policy-making, predicting the poverty rate before the release of Statistics Indonesia (BPS) data is essential. This research compares day and night satellite images to predict per capita expenditure in East Java, Indonesia, which has the highest number of poor people. The satellite images are processed using a transfer learning approach that employs a pretrained Convolutional Neural Network (CNN) model with VGG-16 architecture as a feature extractor. These extracted features are then used as independent variables to predict East Java's per capita expenditure using Support Vector Regression (SVR) with RBF and polynomial kernels. The findings indicate that night images are more reliable than day images, with the best model being a combination of transfer learning and the SVR polynomial kernel using night images. The prediction mapping aligns well with the unmodeled night image, demonstrating the effectiveness of this approach in predicting per capita expenditure.

In this research, the goal was to improve diabetes prediction by combining Multilayer Perceptron Neural Network (MLPNN) with Memetic Algorithm (MA) and Arithmetic Optimization Algorithm (AOA). The method suggested used a preprocessing step to choose a representative subset of attributes from the initial set. Next, the method suggested utilized a combination of the MA and AOA algorithms to optimize feature selection, resulting in a refined dataset that served as input for the Neural Network. Ultimately, the suggested approach utilized the multilayer perceptron neural network (MLPNN) to train the network with hidden layer neurons. The experimental findings indicated a 95% high accuracy rate was achieved. Machine learning classifiers achieved better accuracy compared to classifiers in previous studies, with Decision Tree and Logistic Regression classifiers each reaching 93.57% and 93.33% accuracy, respectively.

This study aims to evaluate an intervention that moved beyond the dimensions of digital information application (DTMs) in science education in higher education. Using an integrated methodology, there is a large and growing body of evidence for the prominence placed on both perceived experience and quality tutors in science courses in educating institutions with broader ambivalence toward limited Digital Literacy. The model and Hypotheses were tested with data collected from 158 participants. Our results suggest that user perceptions about the value of DIE may be contingent on several factors beyond any intrinsic greater experience in learning and teaching Science. The quality of tutors may lead to a greater perceived usefulness regarding the technology. Also, the level of communication flow can affect how much science students are willing to use technology. Academic institutions will have to reassess the utility of digital information technology as an instrument for improving science education. This research focused on educational contexts where DIE profoundly influences teaching and learning science. Further research could focus on other educational fields, math, or language to broaden the understanding of technology's role in diverse educational contexts.