The objective of multidimensional graph fraud detection is to identify fraudulent entities within a graph. Graph Neural Network (GNN) models leverage graph structures to propagate messages from neighbors to target nodes, thereby obtaining precise representations of the target nodes. However, anomaly detection methods based on GNNs face challenges, such as structural inconsistency and over-smoothing, which reduce the suspicion scores of fraudulent nodes and hinder the effectiveness of anomaly detection models. To address these issues, we propose an end-to-end anomaly node detection model named Feature Decomposition Graph Neural Network (FD-GNN). In FD-GNN, a feature amplification module is first employed to enhance the differences between node representations. Then, a feature decomposition module is used to distinguish the natural attributes of nodes from their neighborhood attributes, with constraints applied to both types of attributes. Finally, a representation aggregation module utilizes differential aggregation operations to further distinguish normal nodes from anomaly ones. Experiments conducted on two real-world datasets, Amazon and YelpChi, demonstrate that FD-GNN achieves Area Under Curve (AUC) scores of 94.45% and 91.62%, respectively, outperforming existing multidimensional graph anomaly detection models.

Multilayer networks, as an important branch of network science, have become a powerful tool for revealing and analyzing the internal structures of complex systems. Within these networks, community detection is particularly crucial, as it assists in uncovering hidden patterns within the network. We construct a seed node selection method based on the local structural characteristics of network nodes and, by integrating deep learning methods, establish a local information expansion strategy. This approach effectively identifies and expands community boundaries, developing a novel multilayer network community detection algorithm—the Layered Information Expansion Detection Algorithm (LIEDA). Its exceptional performance has been experimentally verified using multiple real-world datasets. Compared with existing technologies, the LIEDA has considerable accuracy, stability, and adaptability advantages. Compared with various popular benchmark algorithms, the model has substantially improved multiple evaluation metrics across several authoritative public and synthetic datasets.