Search Results (93)

Although there is a growing body of literature on the impact of COVID-19 pandemic, limited evidence exists on the impact of the pandemic on informal female-owned enterprises, and especially those that are located in urban informal settlements. Based on a quantitative survey of 448 vendors selected through stratified random sampling, this study employed a descriptive design to examine the dynamics of women-led informal food vending enterprises across four informal settlements in Nairobi, with particular emphasis on the adverse impacts of the COVID-19 pandemic and the vendors’ coping strategies. Our findings show that women food vendors face numerous challenges which intensified during the pandemic, leading to increased business operation costs, spoilage of perishable products, and oscillating daily sales and profits due to unpredictable market forces. The vendors adopted various strategies to cushion their enterprises and households, including price and stock adjustments; the implementation of hygiene measures; increased use of mobile phones for transactions; reliance on credit, loans, savings, and social networks; temporary business closures; and the relocation of household members to rural areas. These results underscore the critical need for context-specific strategies to support and foster the resilience and sustainability of informal economies during future global pandemics. This is particularly urgent given that very few vendors received some form of institutional support, in addition to having limited access to business loans and training opportunities. Full article

Anomalous transaction behaviors in cryptocurrency markets exhibit high concealment, substantial diversity, and strong cross-modal coupling, making traditional rule-based or single-feature analytical methods insufficient for reliable detection in real-world environments. To address the research focus, a data-centric multimodal anomaly detection framework integrating generative augmentation, latent distribution modeling, and dual-branch real-time detection is proposed. The method employs a generative adversarial network with feature-consistency constraints to mitigate the scarcity of fraudulent samples, and adopts a multi-domain variational modeling strategy to learn the latent distribution of normal behaviors, enabling stable anomaly scoring. By combining the long-range temporal modeling capability of Transformer architectures with the sensitivity of online clustering to local structural deviations, the system dynamically integrates global and local information through an adaptive risk fusion mechanism, thereby enhancing robustness and real-time detection capability. Experimental results demonstrate that the generative augmentation module yields substantial improvements, increasing the recall from $0.421$ to $0.671$ and the F1-score to $0.692$. In anomaly distribution modeling, the multi-domain VAE achieves an area under the curve (AUC) of $0.854$ and an F1-score of $0.660$, significantly outperforming traditional One-Class SVM and autoencoder baselines. Multimodal fusion experiments further verify the complementarity of the dual-branch detection structure, with the adaptive fusion model achieving an AUC of $0.884$, an F1-score of $0.713$, and reducing the false positive rate to $0.087$. Ablation studies show that the complete model surpasses any individual module in terms of precision, recall, and F1-score, confirming the synergistic benefits of its integrated components. Overall, the proposed framework achieves high accuracy and high recall in data-scarce, structurally complex, and latency-sensitive cryptocurrency scenarios, providing a scalable and efficient solution for deploying data-centric artificial intelligence in financial security applications. Full article

Credit card fraud detection remains a major challenge due to extreme class imbalance and evolving attack patterns. This paper proposes a practical hybrid pipeline that combines conditional tabular generative adversarial networks (CTGANs) for targeted minority-class synthesis with Light Gradient Boosting Machine (LightGBM) for classification. Inspired by symmetry principles in machine learning, we leverage the adversarial equilibrium of CTGAN to generate realistic fraudulent transactions that maintain distributional symmetry with real fraud patterns, thereby preserving the structural and statistical balance of the original dataset. Synthetic fraud samples are merged with real data to form augmented training sets that restore the symmetry of class representation. We evaluate Simple Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) classifiers, and a LightGBM model on a public dataset using stratified 5-fold validation and an independent hold-out test set. Models are compared using sensitivity, precision, F-measure(F₁), and area under the precision–recall curve (PR-AUC), which reflects symmetry between detection and false-alarm trade-offs. Results show that CTGAN-based augmentation yields large and consistent gains across architectures. The best-performing configuration, CTGAN + LightGBM, attains sensitivity = 0.986, precision = 0.982, F₁ = 0.984, and PR-AUC = 0.918 on the test data, substantially outperforming non-augmented baselines and recent methods. These findings indicate that conditional synthetic augmentation materially improves the detection of rare fraud modes while preserving low false-alarm rates, demonstrating the value of symmetry-aware data synthesis in classification under imbalance. We discuss generation-quality checks, risk of distributional shift, and deployment considerations. Future work will explore alternative generative models with explicit symmetry constraints and time-aware production evaluation. Full article

Prediction of the Electric Vehicle (EV) charging demand is of great importance to charging stations, especially for newly established charging stations whose demand is difficult to predict due to the absence of past time-series transaction data. This paper develops a deep learning method to fill the literature gap to predict charging demands for the new EV charging stations in the next few days, using a transaction dataset containing over 270 charging stations in Nanjing, eastern China. Specifically, our study introduces the average transactions of neighboring stations as new time-series variables and constructs a Convolutional Neural Network (CNN) model, which is a novel deep learning method. The R-squares of the CNN model achieve an average value of 0.90, which outperforms four time-series prediction models, e.g., the Long Short-Term Memory Network (LSTM) and the Extreme Gradient Boosting (XGBoost). In addition, we visualize the areas with high predicted demand for new charging stations using the trained CNN model and achieve a recommendation accuracy rate of 0.70, providing a reference for EV charging operation companies to find the optimal location of new charging stations. Accurate prediction for new charging stations in this study can provide actionable insights to charging station operators in location selection and create a more favorable EV ecosystem. Full article

Despite significant research interest, the understanding of how to systematically implement Lean practices in supply chains remains limited. Therefore, this study analyzes the impact of blockchain technology on implementing Lean principles within supply chain networks. A theoretical model was developed based on a comprehensive literature review, utilizing innovation diffusion theory, agency theory, and transaction cost economics. The LDA topic modeling, based on big data from the past decade, was employed to explore key areas and essential industry practices related to blockchain technology. By cross-validating big data analysis and survey results, we also developed reliable metrics that can be used to study blockchain utilization in SCM. The hypotheses were empirically tested using survey data from 219 US enterprises that have adopted blockchain technology. The empirical results revealed that blockchain adoption significantly improved Lean management practices within supply chain networks. Furthermore, research has shown that blockchain can significantly enhance operational performance, including cost reduction, quality improvement, delivery capacity, and greater flexibility. These compelling results suggest that blockchain has the potential to serve as a powerful platform for systematically integrating and orchestrating Lean management practices across the entire supply chain network, thereby achieving operational excellence. An in-depth discussion of the study’s practical implications and theoretical contributions is presented. Full article

Urban green spaces (UGS) are pivotal to urban sustainability, yet their morphology—patch size, shape, and configuration—remains insufficiently linked to institutional drivers. We investigate how land tenure strength shapes UGS morphology across 36 cities in nine countries. Using OpenStreetMap data, we delineate UGS and compute landscape metrics (AREA, PARA, SHAPE, FRAC, PAFRAC) via FRAGSTATS; we develop a composite index of land tenure strength capturing ownership, use-right duration, expropriation compensation, and government land governance capacity. Spearman’s rank correlations indicate a scale-dependent coupling: stronger tenure is significantly associated with micro-scale patterns—smaller patch areas and more complex, irregular boundaries—consistent with fragmented ownership and higher transaction costs, whereas macro-scale indicators (e.g., overall green coverage/connectivity) show weaker sensitivity. These findings clarify an institutional pathway through which property rights intensity influences the physical fabric of urban nature. Policy implications are twofold: in high-intensity contexts, flexible instruments (e.g., transferable development rights, negotiated acquisition, ecological compensation) can maintain network connectivity via embedded, fine-grain interventions; in low-intensity contexts, one-off land assembly can efficiently deliver larger, regular green cores. The results provide evidence-based guidance for aligning green infrastructure design with diverse governance regimes and advancing context-sensitive sustainability planning. Full article

Software-Defined Wide-Area Networks (SD-WAN) efficiently manage and route traffic across multiple WAN connections, enhancing the reliability of modern enterprise networks. However, the performance of SD-WANs is largely affected due to malicious activities of unauthorized and faulty nodes. To solve these issues, many machine-learning-based malicious-node-detection techniques have been proposed. However, these techniques are vulnerable to various issues such as low classification accuracy and privacy leakage of network entities. Furthermore, most operations of traditional SD-WANs are dependent on a third-party or a centralized party, which leads to issues such single point of failure, large computational overheads, and performance bottlenecks. To solve the aforementioned issues, we propose a Blockchain Federated-Learning-Enabled Trust Framework for Secure East–West Communication in Multi-Controller SD-WANs (BFL-SDWANTrust). The proposed model ensures local model learning at the edge nodes while utilizing the capabilities of federated learning. In the proposed model, we ensure distributed training without requiring central data aggregation, which preserves the privacy of network entities while simultaneously improving generalization across heterogeneous SD-WAN environments. We also propose a blockchain-based network that validates all network communication and malicious node-detection transactions without the involvement of any third party. We evaluate the performance of our proposed BFL-SDWANTrust on the InSDN dataset and compare its performance with various benchmark malicious-node-detection models. The simulation results show that BFL-SDWANTrust outperforms all benchmark models across various metrics and achieves the highest accuracy (98.8%), precision (98.0%), recall (97.0%), and F1-score (97.7%). Furthermore, our proposed model has the shortest training and testing times of 12 s and 3.1 s, respectively. Full article

The prediction of real estate values is vital for taxation, transactions, mortgages, and urban policy development. Values can be predicted more accurately by statistical or advanced methods together when the size of the data is huge. In metropolitan cities like İstanbul, where size of the real estate data is vast and complex, mass appraisal methods supported by Machine Learning offer a scalable and consistent alternative. This study employs six algorithms: Artificial Neural Network, Extreme Gradient Boosting, K-Nearest Neighbors, Support Vector Regression, Random Forest, and Semi-Log Regression, to estimate the values of real estate on both the Asian and European continent parts of İstanbul. In total, 168,099 residential properties were utilized along with 30 of their features from both sides of the Bosphorus. The results show that RF yielded the best performance in Beşiktaş, while XGBoost performed best in Üsküdar. ANN also produced competitive results, although slightly less accurate than those of XGBoost and RF. In contrast, traditional SVR and SLR models underperformed, especially in terms of R² and RMSE values. With its large-scale dataset, focusing on one of the greatest metropolitan areas, Istanbul, and the usage of multiple ML algorithms, this study stands as a comprehensive and practical contribution to the field of automated real estate valuation. Full article

Accurate evaluation of the operational status of smart meters, as the critical interface between the power grid and its users, is essential for ensuring fairness in power transactions. This highlights the importance of implementing rotation management practices based on meter status. However, the traditional expiration-based rotation method has become inadequate due to the extended service life of modern smart meters, necessitating a shift toward status-driven targeted management. Existing multifactor comprehensive assessment methods often face challenges in balancing accuracy and interpretability. To address these limitations, this study proposes a novel method for analyzing the status of smart meter groups using an interpretable random forest model. The approach incorporates an expert-knowledge-guided grouping assessment strategy, develops a multi-source heterogeneous feature set with strong correlations to meter status, and enhances the random forest model with the SHAP (SHapley Additive exPlanations) interpretability framework. Compared to conventional methods, the proposed approach demonstrates superior efficiency and reliability in predicting the failure rates of smart meter groups within distribution network areas, offering robust support for the maintenance and management of smart meters. Full article

Globally, forest leisure and recreation (FLR) activities are widely recognized not only for their environmental and social benefits but also for their economic contributions. To better understand these economic contributions, it is vital to examine how the regional economic levels of customers vary when consuming FLR. This study aimed to empirically examine whether the regional economic level of residents (i.e., gross regional domestic product; GRDP) is classifiable using FLR expenditure data, and to interpret which variables contribute to its classification. We acquired anonymized credit card transaction data on residents of two regions with different GRDP levels. The data were preprocessed by identifying FLR-related industries and extracting key spending features for classification analysis. Five classification models (e.g., deep neural network (DNN), random forest, extreme gradient boosting, support vector machine, and logistic regression) were applied. Among the models, the DNN model presented the best performance (overall accuracy = 0.73; area under the curve (AUC) = 0.82). SHAP analysis showed that the “FLR industry” variable was most influential in differentiating GRDP levels across all the models. These findings demonstrate that FLR consumption patterns may vary and are interpretable by economic levels, providing an empirical framework for designing regional economic policies. Full article

Practical Byzantine Fault Tolerance (PBFT) has been widely used in consortium blockchain systems; however, it suffers from performance degradation and susceptibility to Byzantine faults in complex environments. To overcome these limitations, this paper proposes RE-BPFT, an enhanced consensus algorithm that integrates a nuanced node credibility model considering direct interactions, indirect reputations, and historical behavior. Additionally, we adopt an optimized ID3 decision-tree method for node classification, dynamically identifying high-performing, trustworthy, ordinary, and malicious nodes based on real-time data. To address issues related to centralization risk in leader selection, we introduce a weighted random primary node election mechanism. We implemented a prototype of the RE-BPFT algorithm in Python and conducted extensive evaluations across diverse network scales and transaction scenarios. Experimental results indicate that RE-BPFT markedly reduces consensus latency and communication costs while achieving higher throughput and better scalability than classical PBFT, RBFT, and PPoR algorithms. Thus, RE-BPFT demonstrates significant advantages for large-scale and high-demand consortium blockchain use cases, particularly in areas like digital traceability and forensic data management. The insights gained from this study offer valuable improvements for ensuring node reliability, consensus performance, and overall system resilience. Full article

This study introduces a data-driven framework for enhancing the sustainable management of fish species in Romania’s Natura 2000 protected areas through ecosystem modeling and association rule mining (ARM). Drawing on seven years of ecological monitoring data for 13 fish species of ecological and socio-economic importance, we apply the FP-Growth algorithm to extract high-confidence co-occurrence patterns among 19 codified conservation measures. By encoding expert habitat assessments into binary transactions, the analysis revealed 44 robust association rules, highlighting interdependent management actions that collectively improve species resilience and habitat conditions. These results provide actionable insights for integrated, evidence-based conservation planning. The approach demonstrates the interpretability, scalability, and practical relevance of ARM in biodiversity management, offering a replicable method for supporting adaptive ecological decision making across complex protected area networks. Full article

This research examined the domain of intelligent transportation systems (ITS) by analyzing the impact of scholarly work and thematic prevalence, as well as focusing attention on vehicles, their technologies, cybersecurity, and related scholarly technologies. This was performed by examining the scientific literature indexed in the Scopus database. This study analysed 2919 documents published between 2018 and 2025. The findings indicated that the highest and most significant journal was derived from IEEE Transactions on Vehicular Technology, with significant standing to the growth of communication and computing on vehicles with edge computing and AI optimization of vehicular systems. In addition, important PST research conferences highlighted the growing interest in academic research in cybersecurity for vehicle networks. Sensor networks, pose forensics, and privacy-preserving communication frameworks were some of the significant contributing fields marking the significance of the interdisciplinary nature of this research. Employing bibliometric analysis, the literature illustrated the multiple channels integrating knowledge creation and innovation in ITS through citation analysis. The outcome suggested an increasingly sophisticated research area, weighing technical progress and increasing concern about security and privacy measures. Further studies must investigate edge computing integrated with AI, advanced privacy-preserving linguistic protocols, and new vehicular network intrusion detection systems. Full article

Artificial Intelligence is a disruptive technology that is revolutionizing the accounting sector, e.g., by reducing costs, detecting fraud, and generating reports. However, the manual maintenance of booking ledgers remains a significant challenge, particularly for small and medium-sized enterprises. The usage of AI technologies in this area is rarely considered in the literature depite a significant interest in using AI for other acounting-related activities. Our study, which was conducted during 2023–2024, utilizes natural language processing and machine learning to construct a predictive model that accurately matches bank transaction statements with accounting records. The study employs Feedforward Neural Networks and Support Vector Machines with various settings and compares their performance with that of previous models embedded in similar predictive tasks. Additionally, as a baseline model, a software called Contofox, a rule-based system capable of classifying accounting records by manually creating rules to match bank statements with accounting records, is used. Furthermore, this study evaluates the business value of the model through an interview with an accounting expert, highlighting the potential benefits of artifacts in enhancing accounting processes. Full article

Partial fingerprint recognition has emerged as a critical method for verifying user authenticity during mobile transactions. As a result, there is a pressing need to develop techniques that effectively and accurately authenticate users, even when the scanner only captures a limited area of the finger. A key challenge in partial fingerprint matching is the inevitable loss of features when a full fingerprint image is reduced to a partial one. To address this, we propose a method that integrates deep learning with feature descriptors for partial fingerprint matching. Specifically, our approach employs a Siamese Network based on a CNN architecture for deep learning, complemented by a SIFT-based feature descriptor to extract minimal yet significant features from the partial fingerprint. The final matching score is determined by combining the outputs from both methods, using a weighted scheme. The experimental results, obtained from varying image sizes, sufficient epochs, and different datasets, indicate that our combined method achieves an Equal Error Rate (EER) of approximately 4% for databases DB1 and DB3 in the FVC2002 dataset. Additionally, validation at FRR@FAR 1/50,000 yields results of about 6.36% and 8.11% for DB1 and DB2, respectively. These findings demonstrate the efficacy of our approach in partial fingerprint recognition. Future work could involve utilizing higher-resolution datasets to capture more detailed fingerprint features, such as pore structures, and exploring alternative deep learning techniques to further streamline the training process. Full article