Jul 2026· 2026 International Conference on Intelligent and Sustainable AI Systems (ICOSAAS)· pp. 838-842· 0 citations· 18 references
Abstract
The typical approach to fraud detection in retail environments relies on rules-based, static detection engines. Such systems perform well against simple, known attack types but struggle against sophisticated, emerging threats and organized financial syndicates. Moreover, rules-based systems produce high false positive rates that seriously degrade the experience of legitimate customers. This paper presents a Graph Artificial Intelligence framework to support high-throughput and streaming retail transactions with built-in explainability. Our approach treats transactions as a dynamic and heterogeneous graph to capture complex relational dependencies and structural anomalies in real time. We validate the framework using a curated subset of the publicly available IEEE-CIS Fraud Detection dataset under a streaming scenario, evaluating both latency and predictive accuracy. Combined with PyTorch Geometric for graph neural network construction and SHAP (Shapley Additive exPlanations) for post-hoc model interpretation, our system delivers high prediction accuracy and human-interpretable decision making. The proposed model shows strong accuracy and recall compared to traditional baselines, while flagged transactions are explained through visual subgraph evidence. This transparency enables fraud analysts to validate alerts efficiently, reducing review time and operational cost in e-commerce environments.
Modern payment fraud increasingly manifests through coordinated activity across cards, devices, merchants, customers, and IP addresses, making transaction-only models insufficient for detecting multi-entity attacks. This paper presents a real-time fraud detection framework based on a heterogeneous temporal graph neural network (HT-GNN) and a two-stage inference architecture that separates offline embedding computation from online transaction scoring. The model represents payment ecosystems as dynamic heterogeneous graphs, uses relation-specific attention and temporal aggregation to learn entity representations, and combines these embeddings with transaction attributes in a lightweight scoring head. On a largescale anonymized dataset containing 847 million transactions and 52 million entities, HT-GNN achieves 65.8 % recall at 0.1 % false positive rate, improving over strong non-graph and graph baselines while maintaining $\mathbf{4 3 ~ m s ~ P 9 5}$ latency. A full runtime analysis shows that the two-stage design constrains end-toend online scoring to 43 ms P95 and offline graph refresh to approximately 28 minutes on four V100 GPUs. Ablation studies show that heterogeneity, device/IP relationships, and temporal attention are the main drivers of performance gains. The results indicate that graph-based models are a practical direction for production-grade fraud detection in payment networks.
Uday Kiran Chilakalapalli, Pranav Sachdev, Brij Mohan et al.· International Conference on...· 0 citations
Financial fraud continues to evolve in sophistication as transaction volumes grow across digital banking, e-commerce, and peer-to-peer payment platforms, rendering traditional rule-based and tabular machine learning detectors increasingly inadequate. This paper proposes a Hybrid Graph Neural Network (GNN) and Explainable Artificial Intelligence (XAI) framework that models financial transactions as a dynamic graph of accounts, merchants, and devices, and learns relational fraud signatures using a combination of GraphSAGE-style neighborhood aggregation and graph attention mechanisms [1-2], [22]. To address the opacity of deep relational models, the framework integrates a post-hoc explainability layer combining SHAP, LIME, and GNNExplainer to generate feature-level and subgraph-level rationales for every fraud alert [4-6]. The proposed model was evaluated on a large-scale, class-imbalanced transaction dataset comprising over 2.1 million transactions and benchmarked against Logistic Regression, Random Forest, XGBoost, standard Graph Convolutional Networks (GCN), and GraphSAGE baselines [3], [17], [21]. Experimental results demonstrate that the proposed hybrid framework achieves 98.6% accuracy, 95.3% precision, 94.1% recall, and a 94.7% F1-score, outperforming the strongest baseline by 3.0 percentage points in F1-score while maintaining an area under the ROC curve (AUC) of 0.989. Ablation experiments confirm that temporal attention and class-imbalance handling each contribute measurable performance gains, and the explainability module improves analyst trust and investigation efficiency by surfacing the top contributing features and the minimal suspicious subgraph behind each decision. The results indicate that combining relational deep learning with transparent explanation mechanisms yields a fraud detection system that is simultaneously more accurate and more auditable than existing approaches, addressing a critical requirement for deployment in regulated financial environments.
Sidharth Shankar, Roopam Bachhil, Praveena Sindagi et al.· International journal of com...· 0 citations
Detecting credit card fraud is a major financial security challenge. This is a result of shifting fraud patterns and extremely unbalanced transaction data. In order to enhance detection performance, this study presents a hybrid fraud detection framework that combines gradient boosting classification with graph-based representation learning. First, transaction features, including the Time attribute considered as a numerical variable, are preprocessed using standard scaling and Borderline-SMOTE to tackle class imbalance. A static k-nearest neighbor similarity graph is created from the scaled feature space. A Temporal Graph Attention Network (TGAT) architecture is used to learn distinct node embeddings from graph relationships, even without explicit temporal encoding. These embeddings are combined with the original transaction features and fed into an XGBoost classifier for final fraud prediction. Optimal decision thresholds are determined through precision-recall analysis. Additionally, SHAP-based explainable artificial intelligence is used to clarify global feature importance and local prediction behavior, which improves model transparency. Experimental evaluation with stratified cross-validation shows that the TGAT-XGBoost hybrid model performs strongly across Accuracy, Precision, Recall, F1-score, ROC-AUC, and PR-AUC metrics. This indicates its effectiveness for reliable and understandable credit card fraud detection.
N. Gopinath, Ampapurapu Devendra Teja, Swathi Koganti· International Conference Com...· 0 citations
Fraud in digital transaction and user-behavior platforms poses escalating threats to financial ecosystems, demanding detection systems that are simultaneously accurate, timely, and interpretable. Although our motivating setting is mobile payment risk control, many public fraud benchmarks contain transaction-like or behavior-fraud graphs rather than strictly mobile payment records. We therefore position the proposed framework as a general graph fraud detection model for heterogeneous interaction graphs, while explicitly analyzing the structural, temporal, and risk-pattern correspondence between payment transactions and review-spam benchmarks. Existing approaches either rely on shallow feature engineering that ignores relational structures among users, merchants, devices, and behavior entities, or employ graph-based models that treat temporal dynamics uniformly and lack mechanisms for explaining their predictions. To address these limitations, we propose TransFraudGNN, a counterfactual-guided dual-timescale graph neural network that jointly models short-term burst behavior and long-term behavioral drift through hierarchical temporal attention. A fraud-aware contrastive pretraining strategy with hard-negative mining improves representation discriminability under severe label scarcity. The explanation module identifies compact prediction-sufficient subgraphs and supports model-level counterfactual reasoning without claiming to recover the real data-generating causal mechanism. Experiments on IEEE-CIS, Amazon, and Yelp demonstrate state-of-the-art AUC-ROC values of 0.9487, 0.9361, and 0.9218, corresponding to relative AUC error reductions of 34.6%, 30.3%, and 26.1% over the strongest baseline.
Minyi Zheng· Journal of King Saud Univers...· 0 citations
It is concluded that each component of a layered fraud system contributes only under specific conditions, and that a plausible rationale from an investigation agent is not evidence of a better decision.
Sophisticated machine learning and deep learning architectures often achieve high predictive performance but remain insufficient for high-stakes domains such as financial fraud detection, where stakeholders require both accuracy and trust in the decision-making process. This study proposes a weighted ensemble framework that strategically combines eXtreme Gradient Boosting (XGBoost) and Graph SAmple and aggreGatE (GraphSAGE) models to optimize the precisionrecall trade-off while maintaining interpretability. Our proposed framework, BoostSAGE, leverages complementary strengths of its constituent models. XGBoost provides high precision (94%) through transaction-level feature analysis, while GraphSAGE ensures superior recall (95%) by capturing structural risk signals across transaction-user-merchant relationships. Empirical results indicate that BoostSAGE outperforms constituent models by synthesizing local and relational feature sets. Most notably, the ensemble delivers a 14.97% improvement in AUCPR over the XGBoost baseline. By leveraging historical fraud aggregation within user-centric connections, the model reduces total False Negatives by 44.8% compared to XGBoost, yielding a final F1score of 92.88%. Our results demonstrate that weighted ensemble framework provides a robust and interpretable approach to fraud detection, balancing operational efficiency with stakeholder trust.
Chaw Thet Zan, Yusuke Maruyama, Felix Beierle· Annual International Compute...· 0 citations