Graph-based methods have proven effective in financial fraud detection by modeling relationships between entities, yet they often fail to leverage the rich textual information present in real-world data. With the ability to understand semantic information, large language models (LLMs) offer a promising solution to enhance fraud detection by incorporating textual data, such as user profiles and transaction descriptions. However, integrating LLMs with graph-based methods introduces two key challenges: (1) the neighborhood camouflage problem, where fraudulent nodes disguise themselves within normal network structures, and (2) the input size constraints of LLMs, making it difficult to process large, complex graphs with extensive textual data. In this paper, we propose a novel framework, Fraud Detection with LLM-enhanced Graph Neural Networks (FLAG), to address these challenges. FLAG integrates LLMs with graph-based fraud detection by introducing two main modules: semantic similarity neighbor sampling, which reduces the input size and further alleviates the influence of camouflaged neighbors by selecting neighbors having high semantic similarity with the target nodes, and LLM-based node enhancement, which extracts discriminative textual features by LLM to enhance node robustness against camouflaged neighbors. To further improve the model, we design a fine-tuning approach that enables the LLM to extract discriminative text more closely related to the node labels, enhancing the model's ability to differentiate between fraudulent and normal nodes. Extensive experiments on public datasets highlight the superiority of FLAG, showing average improvements of 3.14% in F1-macro and 6.97% in AUC. Furthermore, we have deployed FLAG in Alipay's credit risk assessment system and evaluated its performance on a real-world dataset. The results indicate a 0.9% improvement in the KS criterion, further underscoring FLAG's effectiveness.

Graph neural networks (GNNs) have gained popularity in modeling various complex networks, e.g., social network and webpage network. Despite the promising accuracy, the confidences of GNNs are shown to be miscalibrated, indicating limited awareness of prediction uncertainty and harming the reliability of model decisions. Existing calibration methods primarily focus on improving GNN models, e.g., adding regularization during training or introducing temperature scaling after training. In this paper, we argue that the miscalibration of GNNs may stem from the graph data and can be alleviated through topology modification. To support this motivation, we conduct data observations by examining the impacts ofdecisive andhomophilic edges on calibration performance, where decisive edges play a critical role in GNN predictions and homophilic edges connect nodes of the same class. By assigning larger weights to these edges in the adjacency matrix, we observe an improvement in calibration performance without sacrificing classification accuracy. This suggests the potential of a data-centric approach for calibrating GNNs. Motivated by our observations, we propose Data-centric Graph Calibration (DCGC), which uses two edge weighting modules to adjust the input graph for GNN calibration. The first module learns the weights of decisive edges by parameterizing the adjacency matrix and enabling backpropagation of the prediction loss to edge weights. This emphasizes critical edges that fit the prediction needs. The second module computes weights for homophilic edges based on predicted label distributions, assigning larger weights to edges with stronger homophily. These modifications operate at the data level and can be easily integrated with temperature scaling-based methods for better calibration. Experimental results on 8 benchmark datasets demonstrate that DCGC achieves state-of-the-art calibration performance, with an average relative improvement of 36.4% in ECE, while maintaining or even slightly improving classification accuracy. Ablation studies and hyper-parameter analysis further validate the effectiveness and robustness of our proposed method DCGC. Code and data are available at https://github.com/BUPT-GAMMA/DCGC.

This article mainly focuses on the problem of synchronization in finite and fixed time for fully complex-variable delayed neural networks involving discontinuous activations and time-varying delays without dividing the original complex-variable neural networks into two subsystems in the real domain. To avoid the separation method, a complex-valued sign function is proposed and its properties are established. By means of the introduced sign function, two discontinuous control strategies are developed under the quadratic norm and a new norm based on absolute values of real and imaginary parts. By applying nonsmooth analysis and some novel inequality techniques in the complex field, several synchronization criteria and the estimates of the settling time are derived. In particular, under the new norm framework, a unified control strategy is designed and it is revealed that a parameter value in the controller completely decides the networks are synchronized whether in finite time or in fixed time. Finally, some numerical results for an example are provided to support the established theoretical results.

There are few results concerning consensus of multiagent systems (MASs) based on partial differential equations (PDEs), and the problem of how to act boundary control based on distributed measurement on spatial boundary points of MASs has not been solved. This paper addresses boundary control based on distributed measurement for output consensus of leader-following directed MASs modeled by parabolic PDEs. First, a boundary controller acting on spatial boundary points is designed by considering the delivered information produced by agents communicating with neighborhoods. Without considering input constraint, the Lyapunov's direct method is used to obtain a sufficient condition on the existence of the boundary controller to achieve output consensus. The condition is expressed as a form of the feasibility of LMIs. After that, the whole input constraint for MASs is given. And then, one more condition on control gains is obtained to ensure the existence of the boundary controller with input constraint. Finally, one numerical example with two cases illustrates the theoretical analysis results.

This paper analyzes multi-switching network transmission synchronization (MSNTS) problem among three uncertain chaotic systems with unknown parameters. By constructing the effective switching rules, the definition of MSNTS is given and the synchronization schemes are proposed to reach synchronization between any different states of each derive system and any desired states of every respond system by choosing the proper transmission path. Finally, simulation results show the feasibility of research results.

This paper deals with the problem of complex spatio-temporal networks, which is modeled by coupled partial integro-differential equations (PIDEs). A spatial proportional-integral-derivative (SPID) state-feedback controller is studied. With Laypunov direct method, a sufficient condition on synchronization of the complex PIDE network is investigated in terms of linear matrix inequality (LMIs). Finally, a numerical example shows the effectiveness of the proposed results.