The growing reliance on digital technologies among college students heightens their vulnerability to malware, threatening academic performance, privacy, and data security. Despite available protective software, anti-malware adoption remains limited in higher education. This study examines adoption behavior by integrating Technology Threat Avoidance Theory (TTAT) and Mass Interpersonal Persuasion (MIP). Using a quantitative-correlational design with Partial Least Squares Structural Equation Modeling (PLS-SEM), data were collected through a validated survey covering susceptibility, severity, perceived threat, safeguard effectiveness, cost, self-efficacy, persuasion, motivation avoidance, and behavior avoidance. Results show susceptibility and severity shaped perceived threat, which, along with safeguard factors and persuasion, influenced motivation avoidance, strongly predicting behavior avoidance. Findings confirm the TTAT-MIP framework’s applicability and emphasize the need for awareness campaigns combining rational risk communication with peer-driven digital safety practices.
In this paper, the authors designed a method of cross-application sharing for video data within police mobile terminals and implementation in public security converged communication scenario, including a multimedia engine technology based on pipe-filter chain, an application framework based on multimedia engine, and a video data caching mechanism in kernel space based on PF_ring. They performed three cross-application sharing operations on a single video data stream, corresponding respectively to the video stream data processing phase, the video stream data recording phase, and the video stream data transmission phase. By testing three cross-application shared video streams under constant indoor temperature conditions with varying resolutions, they evaluated key performance metrics including MTST, MTLT, and MTPC. The results show that the proposed method effectively addresses the challenge of high-quality, high-speed cross-application sharing, and transmission of video stream data within the resource-constrained police mobile terminals.
Driven by China's “new infrastructure” strategy and “discerning project”, smart community construction has become key to urban digital development. This study designs a video surveillance-based smart community system with a four-tier architecture (perception, network, platform, application layers), integrating AI video analysis, data fusion, blockchain tamper-proof storage, SM4 encryption, and face desensitization to ensure data security—core to the journal's focus. It implements three modules: perimeter intrusion warning (YOLO-based), illegal electric vehicle charging monitoring, and property fee visualization. Empirical results from a 3,000-household pilot (operational since Feb 2024) show 64–66.7% fewer security incidents, 38.9–41.7% shorter property response time, and 26% higher residents' satisfaction. Limitations include algorithm fragility in extreme conditions and multi-device interoperability issues, providing insights for secure smart community development.
This paper evaluates the perceived effectiveness of the security risk management (SRM) programs at a Fortune 500 firm. Layers of management and staff participated in the study. Perceived effectiveness of their SRM programs was based on nine critical success factors (CSFs). Interviews confirmed six initial CSFs (Executive Management Support, Organizational Maturity, Open Communication, Risk Management Stakeholders, Team Member Empowerment, and Holistic View of an Organization) that were extracted from the literature. They were confirmed and synthesized with three additional CSFs (Security Maintenance, Corporate Security Strategy, and Human Resource Development). Implications for SRM are discussed.
The growing reliance on digital technologies among college students heightens their vulnerability to malware, threatening academic performance, privacy, and data security. Despite available protective software, anti-malware adoption remains limited in higher education. This study examines adoption behavior by integrating Technology Threat Avoidance Theory (TTAT) and Mass Interpersonal Persuasion (MIP). Using a quantitative-correlational design with Partial Least Squares Structural Equation Modeling (PLS-SEM), data were collected through a validated survey covering susceptibility, and behavior avoidance. Results show susceptibility and severity shaped perceived threat, which, along with safeguard factors and persuasion, influenced motivation avoidance, strongly predicting behavior avoidance. Findings confirm the TTAT-MIP framework's applicability and emphasize the need for awareness campaigns combining rational risk communication with peer-driven digital safety practices.
Multi-authority attribute-based encryption (MA-ABE) is a generalization of ABE where each party can become an authority. In TCC 2022, Waters et al. constructed an MA-ABE scheme for the class of subset policies from related-trapdoor LWE. However, their security reduction imposes significant overhead on the security and efficiency of their scheme. In this paper, the authors propose some improvements to this MA-ABE scheme. Firstly, they optimize the size of the public key and simplify the security proof to make the security reduction more compact. To further improve the encryption efficiency, they construct a new scheme over rings and reduce it directly to ring LWE. To generate a user’s secret key in our scheme, they extend the Gaussian sampling algorithm over ideal lattices. Finally, they streamline the statement and proof of the Leftover Hash Lemma over rings to make the security proof of our MA-ABE scheme more concise.
Protection motivation theory (PMT) has long explained information security behavior as a rational evaluation of threats and coping mechanisms. However, existing models provide limited insight into how organisational conditions translate into information security experts’ personal responsibility for protecting information assets. This study addresses this gap by integrating psychological ownership (PO) into PMT and reconceptualising protection motivation as an ownership-driven process. Using survey data from 1,254 information security professionals in Chinese IT organisations and structural equation modelling, the results show that PO is the strongest predictor of protective intention. PO significantly strengthens threat appraisal and self-efficacy, whereas response efficacy and response cost do not significantly influence intention. These findings suggest that security behavior is driven less by calculative threat evaluation than by the internalisation of organisational responsibility, offering a psychologically grounded perspective on sustainable information security engagement.
High real-time mobile video stream cross-domain secure transmission and sharing has become a research focus in the field of high-performance and secure communication. In this study, a more efficient method based on combining a zero-copy model with a one-way light transmission mechanisms is proposed. The method contains an effective integrated framework and layered design structure. The latest PF_RING (9.0.0) zero-copy model, memory routing table, multi-queue multi-threading, and related technologies are comprehensively combined to ensure the high-performance cross-domain secure transmission and sharing of video streams. At the same time, to ensure the safe transmission of shared mobile video data between different network domains, both a mobile video surveillance data transmission model and an anti-storage-based covert channel model are proposed. The experimental results show that the performance of the proposed method has increased by at least 10% in throughput, central processing unit utilization, and bit error rate.
With the extensive proliferation of location-based services, protecting user trajectory privacy against continuous query attacks has become a critical challenge. Existing protection mechanisms often suffer from a rigid trade-off between privacy strength and service quality. To bridge this gap, this study proposes a unified demand-aware trajectory privacy protection framework. First, a fake trajectory generation algorithm is developed that ensures to resist advanced inference attacks. Second, a maximizing demand request algorithm is introduced to resolve conflicts between privacy demands and sparse historical data. Finally, two anonymous zone minimization strategies are implemented. Experimental results using real-world mobility generators demonstrate that the proposed framework improves the anonymous service success rate by more than 13% over baseline location privacy-preserving algorithms while maintaining a smaller anonymous area, balancing privacy and utility
Federated learning (FL) enables collaborative training without sharing raw data, but standard FL exposes client updates and burdens resource-constrained IoT devices. The authors propose AdaptiveCKKS, an FL framework combining aggregation-only CKKS encryption with index-free block sparsification and stochastic quantization. A lightweight controller adaptively selects compression ratio and quantization per device/round based on on-device calibration of bandwidth, CPU, and encryption cost, while CKKS contexts are fixed at enrollment. The server performs ciphertext-only additions, decrypting only the aggregate each round. On BoT-IoT and ToN-IoT datasets, AdaptiveCKKS improves accuracy by 3.2–3.8% over FL and fixed-HE, reduces per-round communication by ~45% and average power by ~39%, and increases resistance to membership inference and gradient inversion attacks. Results are averaged over 10 runs with 95% confidence intervals, and all artifacts are released for reproducibility.
The growing sophistication of cyber threats demands adaptive security mechanisms beyond traditional Intrusion Detection Systems (IDS). This paper explores integrating Machine Learning (ML) and Explainable Artificial Intelligence (XAI) to enhance Network Intrusion Detection Systems (NIDS). Using the CICIDS2017 dataset, the authors evaluate ML models including Convolutional Neural Networks (CNN), Random Forest, and XGBoost, balancing detection performance with interpretability. Results show XGBoost achieves the highest accuracy with minimal misclassifications, underscoring its robustness for intrusion detection. To address the black-box challenge of deep learning, SHapley Additive exPlanations (SHAP) is applied to interpret predictions. Key features such as Destination Port, Flow Duration, and Packet Length emerged as critical, improving trust, reducing false positives, and aiding investigation. The authors highlight the necessity of coupling high-performing ML with XAI frameworks for transparency. Finally, challenges in scalability, robustness, and dataset generalizability are discussed.
Existing multi-domain personalized local differential privacy (MDPLDP) mechanisms, which extend attribute domains by introducing fake values, often fail to provide adequate personalized privacy protection and limit utility in frequency estimation. To address these limitations, the authors propose two novel MDPLDP mechanisms that construct multiple domains by partitioning real attribute values, support cross-domain aggregation, and flexibly accommodate diverse privacy requirements and budgets. The methods further extend to multi-dimensional frequency estimation, catering to complex user privacy preferences. Theoretical analysis and experimental results demonstrate that our mechanisms achieve substantially lower estimation error and communication overhead, while delivering over 20% average utility improvement compared to state-of-the-art methods in both single- and multi-dimensional settings.
Digital art growth brings copyright challenges: slow verification, tampering risks, and privacy leaks. This study proposes a blockchain-based cryptographic framework for secure, private, and traceable management. The layered design uses AES for content encryption, ECC for key security, and blockchain for immutable metadata storage, ensuring integrity and authenticity. Smart contracts enable automated access control with pseudonymized identities to protect user privacy. Tested on 2D, 3D, and dynamic artworks, the system outperforms traditional DRM: verification latency drops over 50%, tamper detection exceeds 95%, and CPU/memory usage stays low. It supports scalable, real-time operations and provides an auditable, trust-aware environment for content lifecycle management. Though interoperability and new formats remain challenges, the framework meets technical, organizational, and regulatory needs in information security. This work advances integrated, privacy-preserving solutions for sustainable digital rights ecosystems.
This paper evaluates the perceived effectiveness of the security risk management (SRM) programs at two Fortune 500 firms using qualitative and quantitative methods. Layers of management and staff from both firms participated in the study. Perceived effectiveness of their SRM programs was based on nine critical success factors (CSFs). Six initial critical success factors (CSFs): executive management support, organizational maturity, open communication, risk management stakeholders, team member empowerment, and holistic view of an organization were extracted from organizational role theory. They were confirmed and synthesized with three additional CSFs (security maintenance, corporate security strategy, and human resource development). A survey based on the CSFs was implemented at the two firms. Although both firms are Fortune 500 technology companies, their perceptions of current perceived SRM effectiveness differ significantly.
The authors assess whether state public-safety spending relates to reported cybercrime in a single-year (2023) U.S. state-level analysis (50 states + DC). IC3/FTC outcomes (rates and losses) are linked to police/corrections spending and pre-specified socio-economic/infrastructure covariates. Models use population-weighted least squares for rate outcomes, HC3 robust SEs, ln(loss+1) for dollar outcomes, and Benjamini-Hochberg FDR (q=.10). Zero-order correlations between spending and reporting measures are positive; after adjustment, coefficients attenuate toward zero and are not FDR-significant. Sensitivity checks (excluding DC; omitting influential states; police vs. corrections; IC3-only; adult denominators) are consistent. Findings are associational, consistent with visibility/ detection or demand-driven dynamics rather than cross-sectional deterrence. Practice should pair investigative capacity with standardized reporting and prevention; causal evaluation requires longitudinal/quasi-experimental designs.
In recent years, the development of intelligent building technology has not only improved the quality and efficiency of construction projects, but also triggered complex legal compliance risks. This paper aims to build a dynamic compliance risk control model suitable for intelligent construction environment to deal with these complex legal risks. Based on the logical framework of “multi-agent-multi-process-multi-data source”, the model emphasizes the triple closed loop of information flow, decision flow, and compliance control flow and pays attention to risk feedback and rapid adjustment mechanism. Through theoretical analysis, case verification, model construction, and empirical testing, this paper uses automatic early warning system and real-time data flow to improve the real-time, adaptive, and systematic compliance management, thus effectively protecting data security and personal privacy and reducing legal risks. This provides an operable and extensible compliance risk management system for construction enterprises, which is helpful to improve the project management level.
The authors assess whether state public-safety spending relates to reported cybercrime in a single-year (2023) U.S. state-level analysis (50 states + DC). IC3/FTC outcomes (rates and losses) are linked to police/corrections spending and pre-specified socio-economic/infrastructure covariates. Models use population-weighted least squares for rate outcomes, HC3 robust SEs, ln(loss+1) for dollar outcomes, and Benjamini–Hochberg FDR (q=.10). Zero-order correlations between spending and reporting measures are positive; after adjustment, coefficients attenuate toward zero and are not FDR-significant. Sensitivity checks (excluding DC; omitting influential states; police vs. corrections; IC3-only; adult denominators) are consistent. Findings are associational, consistent with visibility/detection or demand-driven dynamics rather than cross-sectional deterrence. Practice should pair investigative capacity with standardized reporting and prevention; causal evaluation requires longitudinal/quasi-experimental designs.
This study explores the relatively under-researched area of comparing data privacy regulations and best practices across different countries, with a focus on the gaming industry. It provides an overview of general data privacy principles and existing global regulations, analyzing how gaming operators leverage personal data for competitive advantage. Specifically, the research examines the data privacy approaches and regulatory requirements in Singapore, Macau, and Japan, highlighting the cultural and historical contexts influencing these regulations. Through a comparative analysis, the article discusses the compliance needs for gaming operators in these jurisdictions.
The rapid and substantial evolution of technology and multimedia applications demands significant resources to ensure enhanced performance. As a solution, designers are increasingly turning to manycore architectures with 3D-NoCs for communication. However, technologies such as vertical nano-links and the short distances between stacked layers, along with limitations in buffer sizes, contribute to link failures and heat dissipation between layers. These factors degrade NoC performance and give rise to issues such as congestion and link failures. The authors propose a fault-tolerant and congestion control routing algorithm in 3D-Network-on-Chip, inspired by bird flocking behavior. The approach is designed to respect the resource constraints of NoCs in terms of processing, buffer sizes, and power consumption. Simulation results demonstrate that the proposal achieves superior performance compared to the classic XYZ routing algorithm in term of delay and throughput.
Facial expression recognition (FER), as an important branch of computer vision, has made significant progress in recent years thanks to the development of deep learning technology. This article proposes a support vector machine (SVM) facial expression recognition algorithm based on convolutional neural network (CNN) optimization, aiming to improve recognition accuracy and robustness. This method utilizes the advantages of traditional machine learning such as SVM and the automatic feature extraction characteristics of deep learning, proposes a new feature extraction and classification model, and further improves the model performance through data augmentation, hyperparameter optimization, and other means. Experimental verification shows that the algorithm exhibits good recognition accuracy and robustness on multiple publicly available datasets. This study provides new ideas for improving the performance of facial expression recognition systems, which is of great significance for promoting progress in this field.
