Expert Systems with Applications最新文献

{"title":"Identification of critical nodes by fusing propagation probabilities and entropy in binary networks","authors":"Lintao Zhang ,&nbsp;Jianing Zhang ,&nbsp;Rong Yan ,&nbsp;Guoqin Yu","doi":"10.1016/j.eswa.2025.129861","DOIUrl":"10.1016/j.eswa.2025.129861","url":null,"abstract":"<div><div>Identification of critical nodes is crucial for effectively allocating resources and prioritizing tasks in complex networks, which significantly enhances the stability and the efficiency of networks in real-world environments. Generally, existing studies primarily focus on extracting multiple different influential factors from network topology, but they have to face accuracy limitations due to high computational complexity, overlapping influence ranges, and information loss. Inspired by information entropy, in this paper, we explore to identify critical node in complex networks from the perspective of inter-node propagation probabilities. We introduce an innovative critical node ranking algorithm, named MNIE (Mixed Node Information Entropy). MNIE initially segments the node influence within the network topology by distinguishing between global and local effects so as to integrate a more comprehensive topological features set. Then, we refine the connection probability calculation and integrate the features derived from the network structural topology with the probabilities of information transmission (infection rates) among the nodes. Experimental results on 9 real-world networks and 4 synthetic datasets indicate that MNIE enhances the identification of critical nodes and accomplishes better than state-of-the-art methods on monotonicity and accuracy.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129861"},"PeriodicalIF":7.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221474","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CoSemiGNN: Blockchain fraud detection with dynamic graph neural networks based on co-association of semi-supervised","authors":"Yulong Wang ,&nbsp;Qingxiao Zheng ,&nbsp;Xuedong Li ,&nbsp;Lingfeng Wang ,&nbsp;Ling Lin","doi":"10.1016/j.eswa.2025.129853","DOIUrl":"10.1016/j.eswa.2025.129853","url":null,"abstract":"<div><div>With the development of blockchain technology, the increasing number of cyber frauds has caused huge economic losses, prompting more and more researchers to focus on how to effectively detect criminal activities in the blockchain transaction environment. Currently, graph neural network (GNN)-based methods have made significant progress in the field of blockchain illegal transaction detection due to their advantages in extracting graph structure features. However, existing illegal transaction pattern detection methods usually rely on historical labeled data. In the blockchain transaction environment, transaction data changes over time, and it is often difficult to obtain transaction labels. As a result, the performance of these methods is often unsatisfactory when faced with newly distributed transaction data. To address this challenge, this paper proposes a dynamic graph neural network based on co-association of semi-supervised (CoSemiGNN) for more efficiently identifying illegal transactions in blockchain environments under conditions of dynamically changing transaction data. The model combines semi-supervised learning with a dynamic graph neural network, enabling it to effectively identify novel illegal transaction patterns from unlabeled data and adapt to the evolving blockchain network environment. Specifically, CoSemiGNN captures features of novel transactions by integrating semi supervised learning results. It utilizes co-occurrence relations of edges and co-occurrence feature aggregation of nodes to skillfully integrate semi-supervised methods into feature extraction of transaction graphs, enabling the model to extract novel illegal transaction patterns from unlabeled data. In addition, the model utilizes self attention recurrent neural networks (RNNs) to capture temporal information in transactions, ensuring the dynamics of CoSemiGNN. Finally, we theoretically analyze the model, and experiments on a real Bitcoin transaction dataset demonstrate that CoSemiGNN outperforms existing methods by as much as 30 % in terms of F1 scores for detecting illegal transactions when the transaction data undergoes distributional migration. This research compensates the problem that existing methods ignore the distributional changes of blockchain transaction data, and provides a new perspective and an effective solution for blockchain illegal transaction detection.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129853"},"PeriodicalIF":7.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IMPACT-Net: An integrated multi-scale and computation-efficient timely network for surface defect detection in industrial embedded systems","authors":"Ruiqi Wu ,&nbsp;Yong Zhang ,&nbsp;Rukai Lan ,&nbsp;Lei Zhou","doi":"10.1016/j.eswa.2025.129867","DOIUrl":"10.1016/j.eswa.2025.129867","url":null,"abstract":"<div><div>Automated defect detection is crucial in industrial production, with typical scenarios in steel, automotive, and new energy manufacturing. Although deep learning-based defect detection methods have achieved significant progress, challenges such as limited detection accuracy for low-contrast defects and difficulties in efficient inference still persist. To address these challenges, this paper proposes an integrated multi-scale and computation-efficient timely network (IMPACT-Net) for surface defect detection. Firstly, a Precision-Enhanced Feature Pyramid Network (PE-FPN) is designed to improve the detection performance for low-contrast and fine defects by enhancing multi-scale feature fusion. Secondly, an Adaptive Normalized Wasserstein Distance Loss (ANWD-Loss) is proposed to optimize bounding box localization accuracy and enhance robustness. Finally, by employing Progressive Block-Freezing Architecture Search (PBF-AS) and a ZYNQ-based acceleration platform, computational complexity is significantly reduced, and efficient inference is achieved under low-power conditions. Experimental results show that the proposed IMPACT-Net achieves an mAP50 of 78.9 % on the NEU-DET dataset with 2.3 ms inference time and 71.4 % mAP50 on the GC10-DET dataset with 2.5 ms inference time, demonstrating a good balance between detection accuracy and real-time performance that is well suited for resource-constrained embedded industrial environments.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129867"},"PeriodicalIF":7.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221877","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Short-term air quality prediction using a multi-scale attention fusion model with 3DIGAT-CBAM-BiLSTM based on spatio-temporal correlation","authors":"Liangqiong Zhu ,&nbsp;Liren Chen ,&nbsp;Huayou Chen","doi":"10.1016/j.eswa.2025.129856","DOIUrl":"10.1016/j.eswa.2025.129856","url":null,"abstract":"<div><div>Air Quality Index (AQI) prediction is crucial for environmental management and public health. However, most existing studies focus on single site modeling, neglecting the complex spatial correlations of meteorological factors and air pollutants. Therefore, a multi-scale spatio-temporal prediction model, 3DIGAT-CBAM-BiLSTM, is proposed to fully capture the spatio-temporal evolution characteristics of AQI. To reduce the interference of redundant information, the Maximum Information Coefficient and Dynamic Time Series Trend Correlation Method are employed to select the neighboring sites and influencing factors that are highly correlated with the AQI of the target site. The original air quality data is decomposed and reconstructed into high-frequency, low-frequency, and trend-term subsequences using Multivariate Variational Mode Decomposition and Sample Entropy to enhance prediction accuracy. To forecast the three-dimensional spatial tensors of these reconstructed subsequences based on time steps, monitoring sites, and influencing factors, we propose the 3DIGAT-CBAM-BiLSTM model. The spatial dependencies between sites are effectively captured by the Improved Graph Attention Network, which constructs a graph adjacency matrix based on MIC and geographic distance. Meanwhile, the Convolutional Block Attention Mechanism enhances the focus on important sites and features by combining channel and spatial attention. Furthermore, the Bidirectional Long Short-Term Memory network extracts global temporal patterns. The experimental results on the Beijing dataset show that the proposed model achieves a relative reduction of 8.53 % in RMSE and 5.83 % in MAE compared with the optimal baseline model, demonstrating clear performance improvements and offering a novel approach for modeling complex spatio-temporal data.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129856"},"PeriodicalIF":7.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SmartScope: Smart contract vulnerability detection via heterogeneous graph embedding with local semantic enhancement","authors":"Zhaoyi Meng,&nbsp;Zexin Zhang,&nbsp;Wansen Wang,&nbsp;Jie Cui,&nbsp;Hong Zhong","doi":"10.1016/j.eswa.2025.129857","DOIUrl":"10.1016/j.eswa.2025.129857","url":null,"abstract":"<div><div>Smart contracts are integral to blockchain ecosystems, yet their security remains a critical concern due to the prevalence of exploitable vulnerabilities. Existing conventional and deep learning-based vulnerability detection methods often struggle to capture the fine-grained semantics and heterogeneous structural dependencies essential for accurate analysis. We propose and implement <span>SmartScope</span>, a novel technique for smart contract vulnerability detection that leverages heterogeneous graph embedding with local semantic enhancement. Specifically, <span>SmartScope</span> constructs a semantically rich contract graph that depicts control-flow, data-flow, and fallback relations among critical code elements. To guide the graph learning process, we empirically assign various importance coefficients to vulnerability-relevant subgraphs, thereby enhancing the detection model’s focus on semantically critical regions. The heterogeneous graph transformer is then employed to generate context-aware node representations, which are then passed to an MLP-based detector for vulnerability classification. To the best of our knowledge, this is the first method that structurally encodes domain knowledge into the heterogeneous graph learning for achieving effective smart contract analysis. Experimental results demonstrate that <span>SmartScope</span> outperforms 10 representative conventional and deep learning-based baselines on over 5K smart contracts. The evaluation spans multiple vulnerability types, including reentrancy, timestamp dependence, and infinite loops, highlighting the effectiveness and robustness of our work.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129857"},"PeriodicalIF":7.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Disruption-responsive berth allocation and quay crane scheduling with inter-terminal collaboration","authors":"Hongxing Zheng ,&nbsp;Zhaoyang Wang ,&nbsp;Lingxiao Wu","doi":"10.1016/j.eswa.2025.129776","DOIUrl":"10.1016/j.eswa.2025.129776","url":null,"abstract":"<div><div>Container terminal operations frequently encounter disruptions, including delays, extended handling times, and unscheduled vessel arrivals, all of which necessitate intelligent rescheduling strategies to maintain operational efficiency. This study investigates the integrated problem of disruption-responsive berth allocation and quay crane (QC) scheduling, explicitly considering vessel gathering status and incorporating inter-terminal shifting (ITS) and reassignment to terminals different from its originally designated one (RT) as adaptive response strategies to mitigate these disruptions. A rescheduling model is developed to minimize associated costs. To efficiently solve large-scale problems, an adaptive large neighborhood search (ALNS)-based heuristic is proposed. The effectiveness of the proposed scheme is validated through comparative experiments involving three alternative schemes, highlighting its superior performance. Furthermore, algorithm comparison experiments are conducted to verify the robustness of parameter settings. Computational results demonstrate that the proposed model and algorithm achieve high efficiency and solution quality. Additionally, sensitivity analysis reveals that neglecting vessel gathering status leads to substantial cost increases, particularly in large-scale operations. The integration of ITS and RT proves to be an effective strategy for mitigating disruptions, enhancing scheduling flexibility, and improving operational performance.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129776"},"PeriodicalIF":7.5,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Frequency-decomposed attention joint optimization network for image compressive sensing","authors":"Zhifu Tian,&nbsp;Tao Hu,&nbsp;Di Wu,&nbsp;Shu Wang,&nbsp;Tingli Li,&nbsp;Ming Zhang","doi":"10.1016/j.eswa.2025.129866","DOIUrl":"10.1016/j.eswa.2025.129866","url":null,"abstract":"<div><div>Network approaches for image compressive sensing (ICS) have garnered significant attention due to their high efficiency and fidelity in image reconstruction. Reconstructing complex image textures from highly compressed measurements has been a longstanding goal of ICS, yet existing methods often struggle to varying degrees with the restoration of low-frequency (LF) textures and high-frequency (HF) details, which potentially limits the quality of the reconstructed image. In this paper, we propose a Frequency-decomposed Attention Joint Optimization Network (FAJO-Net) for ICS, which is capable of enhancing the attention to LF and HF components of images. Specifically, we introduce a frequency-decomposed sparse prior and coupling fidelity constraints, and incorporate a tri-optimization network framework for full, low, and high-frequency (FLH) features, where each component is optimized using an optimization-unfolded multi-scale network (OM-Net), inclusive of Principal Component Augmented Gradient Descent Module (PCAGDM) and U-shaped Proximal Mapping Module (UPMM). The PCAGDM optimizes the FLH features efficiently by supplementing the optimization of the minimum dimension principal component augmented features while optimizing the principal component features. The UPMM is able to perform multi-scale proximal mapping for all FLH features. Finally, we design a Frequency-decomposed Interaction Attention Module (FIAM) to enhance the fusion of FLH features, particularly the HF and LF components related to the full-frequency features, while reducing the impact of unnecessary features introduced by frequency decomposition. Extensive experiments demonstrate that our proposed FAJO-Net surpasses the state-of-the-art ICS networks in terms of image fidelity and visual effect, and validates that the proposed FAJO-Net framework can help enhance the image reconstruction capabilities of the vast majority of existing ICS networks, further unlocking the potential for high-fidelity restoration in ICS. Code is available at <span><span>https://github.com/giant-pandada/FAJO-Net</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129866"},"PeriodicalIF":7.5,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222083","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"GraphShield: Advanced dynamic graph-based malware detection using graph neural networks","authors":"Eslam Amer ,&nbsp;Shaker El-Sappagh ,&nbsp;Tamer Abuhamad ,&nbsp;Bander Ali Saleh Al-Rimy ,&nbsp;Alaa Mohasseb","doi":"10.1016/j.eswa.2025.129812","DOIUrl":"10.1016/j.eswa.2025.129812","url":null,"abstract":"<div><div>The rising complexity of modern malware-such as polymorphic, fileless, and sandbox-aware variants-has severely diminished the reliability of conventional detection techniques. Models based on sequential data frequently miss intricate behavioral patterns and long-range dependencies, resulting in poor accuracy and limited adaptability to new threats. This paper introduces GraphShield, a graph-centric behavioral detection framework that identifies malware with high precision by analyzing dynamic API call sequences. GraphShield converts raw API calls into temporal graphs, applies semantic vectorization, and leverages attention mechanisms to extract both localized activity and extended behavioral correlations, directly addressing the weaknesses of earlier systems. We design and assess multiple Graph Neural Network (GNN) variants, including Graph Convolutional Networks (GCNs), Graph Attention Networks (GATs), Graph Isomorphism Networks (GINs), and Transformer-based architectures combining convolutional, recurrent, and autoencoding layers. These models capture structural and temporal traits of execution traces using both classification-only and combined classification-reconstruction strategies. To enhance transparency, we incorporate GNN interpretation tools that isolate key API call subgraphs and critical decision pathways, making detection outcomes explainable for analysts. GraphShield is trained on 300,000 balanced instances and tested on a separate 200,000-sample holdout set, achieving over 58 % improvement in accuracy over advanced sequence-driven deep learning models while maintaining a false positive rate under 1 %. Key features include BERT-based API call grouping for reducing dimensionality and a Markov-inspired graph stabilization method for managing graphs of variable length. Our top models attain a 99.5 % F1-score on the test set. GraphShield aligns recent graph learning techniques with operational cybersecurity needs, delivering accurate detection and clear, interpretable results.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129812"},"PeriodicalIF":7.5,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222088","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"RAFN: A risk-aware feature network for identifying risk factors in supply chain finance","authors":"Yang Zhang ,&nbsp;Yating Zhao ,&nbsp;Wenjuan Lian ,&nbsp;Bin Jia","doi":"10.1016/j.eswa.2025.129874","DOIUrl":"10.1016/j.eswa.2025.129874","url":null,"abstract":"<div><div>As supply chain finance businesses expand, traditional risk assessment systems, which rely heavily on manual processes and static rule-based frameworks, are increasingly unable to keep up with the complexity and dynamism of modern risk patterns. This often leads to delayed responses and inefficiencies in risk management. To address key challenges such as difficulties in integrating heterogeneous data, low detection rates for hidden risks, and limited ability to capture dynamic risk patterns, this paper introduces a novel Risk-Aware Feature Network (RAFN) driven by an adaptive attention mechanism. The RAFN model is designed with a dual-channel architecture to process numerical and categorical data separately, employs gated linear units to dynamically merge heterogeneous data streams, and incorporates a multi-head attention mechanism with dynamic coefficients to focus on risk-sensitive features adaptively. Experiments conducted on both public and proprietary datasets show that RAFN outperforms mainstream algorithms, achieving a 1.73%-5.81% improvement in accuracy, recall, and F1-score, while maintaining a strong balance between specificity and recall. Furthermore, this study proposes a closed-loop risk management framework based on RAFN, which integrates “smart contract triggering, off-chain model evaluation, and on-chain consensus validation.” This approach offers an efficient technical solution to break down data silos and enhance the precision of risk identification in supply chain finance, paving the way for more effective and reliable risk control systems.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129874"},"PeriodicalIF":7.5,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159216","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Markov-based continuous learning with diversion of data distribution direction for streaming data in limited memory","authors":"Peemapat Wongsriphisant ,&nbsp;Kitiporn Plaimas ,&nbsp;Chidchanok Lursinsap","doi":"10.1016/j.eswa.2025.129818","DOIUrl":"10.1016/j.eswa.2025.129818","url":null,"abstract":"<div><div>Traditional online classifiers often require accumulating past data, leading to uncontrollable memory usage and learning times. The ideal solution is a Markov-based continuous learning approach, where a model updates using only its current state and new data. While one-pass learning with hyper-ellipsoids aligns with this principle, a critical weakness persists: classification ambiguity for data points within the overlap region where ellipsoids from different classes intersect. To solve this, this paper proposes the Diversion of Data Distribution Direction (D⁴), a new method that implements this Markov-based approach while specifically targeting the ambiguity problem. D⁴ introduces two novel mechanisms: a new adaptive width adjustment to prevent over-adjusted ellipsoid boundaries and a distribution diversion technique that resolves ambiguity by projecting data into an optimally selected subspace. The proposed D⁴ method was evaluated against seven state-of-the-art online classifiers across nine benchmark datasets, having 2011 to 567,498 samples. It achieved the highest accuracy and macro F1-score on six datasets while proving to be the most computationally efficient and generating the most compact models.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129818"},"PeriodicalIF":7.5,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221874","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}