Search Results (10)

Fraud detection in large-scale graphs presents significant challenges, especially in heterophilic graphs where linked nodes often belong to dissimilar classes or exhibit contrasting attributes. These asymmetric interactions, combined with class imbalance and limited labeled data, make it difficult to fully leverage node labels in semi-supervised learning frameworks. This study aims to address these challenges by proposing a novel framework, Spectrum-Constrained and Skip-Enhanced Graph Fraud Detection (SCSE-GFD), designed specifically for fraud detection in heterophilic graphs. The primary objective is to enhance fraud detection performance while maintaining computational efficiency. SCSE-GFD integrates several key components to improve performance. It employs adaptive polynomial convolution to capture multi-frequency signals and utilizes relation-specific spectral filtering to accommodate both homophilic and heterophilic structures. Additionally, a relation-aware mechanism is incorporated to differentiate between edge types, which enhances feature propagation across diverse graph connections. To address the issue of over-smoothing, skip connections are used to preserve both low- and high-level node representations. Furthermore, supervised edge classification is used to improve the structural understanding of the graph. Extensive experiments on real-world datasets, including Amazon and YelpChi, demonstrate SCSE-GFD’s effectiveness. The framework achieved state-of-the-art AUC scores of 96.21% on Amazon and 90.58% on YelpChi, significantly outperforming existing models. These results validate SCSE-GFD’s ability to improve fraud detection accuracy while maintaining efficiency. Full article

Graph Neural Network (GNN) is an effective model for processing graph-structured data. Most GNNs are designed to solve homophilic graphs, where all nodes belong to the same category. However, graph data in real-world applications are mostly heterophilic, and homophilic GNNs cannot handle them well. To address this, we propose a novel hybrid-learning framework based on Node Homophily and Spectral Heterophily (NHSH) for node classification in graph networks. NHSH is designed to achieve state-of-the-art or superior performance on both homophilic and heterophilic graphs. It includes three core modules: homophilic node extraction (HNE), heterophilic spectrum extraction (HSE) and node feature fusion (NFF). More specifically, HNE identifies symmetric neighborhoods of nodes with the same category, extracting local features that reflect these symmetrical structures. Then, HSE uses filters to analyze the high and low-frequency information of nodes in the graph and extract the global features of the nodes. Finally, NFF fuses the above two node features to obtain the final node features in graphs. Moreover, an elaborate loss function drives the network to preserve critical symmetries and structural patterns in the graph. Experiments on eight benchmark datasets validate that NHSH performs comparably or better than existing methods across diverse graph types. Full article

Tree-like structures, characterized by hierarchical relationships and power-law distributions, are prevalent in a multitude of real-world networks, ranging from social networks to citation networks and protein–protein interaction networks. Recently, there has been significant interest in utilizing hyperbolic space to model these structures, owing to its capability to represent them with diminished distortions compared to flat Euclidean space. However, real-world networks often display a blend of flat, tree-like, and circular substructures, resulting in heterophily. To address this diversity of substructures, this study aims to investigate the reconstruction of graph neural networks on the symmetric manifold, which offers a comprehensive geometric space for more effective modeling of tree-like heterophily. To achieve this objective, we propose a graph convolutional neural network operating on the symmetric positive-definite matrix manifold, leveraging Riemannian metrics to facilitate the scheme of information propagation. Extensive experiments conducted on semi-supervised node classification tasks validate the superiority of the proposed approach, demonstrating that it outperforms comparative models based on Euclidean and hyperbolic geometries. Full article

The graph neural network (GNN) is a type of powerful deep learning model used to process graph data consisting of nodes and edges. Many studies of GNNs have modeled the relationships between the edges and labels of nodes only by homophily/heterophily, where most/few nodes with the same label tend to have an edge between each other. However, this modeling method cannot describe the multiconnection mode on graphs where homophily can coexist with heterophily. In this work, we propose a transition matrix to describe the relationships between edges and labels at the class level. Through this transition matrix, we constructed a more interpretable GNN in a neighbor-predicting manner, measured the information that the edges can provide for the node classification task, and proposed a method to test whether the labels match the edges. The results show the improvement of the proposed method against state-of-the-art (SOTA) GNNs. We also obtain the following two results: (1) the poor performance of GNNs is highly relevant to the information of edges instead of heterophily, which is always considered the main factor resulting in the decline in performance; and (2) most benchmark heterophilic datasets exhibit the label-edge mismatch problem, leading them to become intractable Full article

Graph neural networks (GNNs) have been widely applied for hyperspectral image (HSI) classification, due to their impressive representation ability. It is well-known that typical GNNs and their variants work under the assumption of homophily, while most existing GNN-based HSI classification methods neglect the heterophily that is widely present in the constructed graph structure. To deal with this problem, a homophily-guided Bi-Kernel Graph Neural Network (BKGNN) is developed for HSI classification. In the proposed BKGNN, we estimate the homophily between node pairs according to a learnable homophily degree matrix, which is then applied to change the propagation mechanism by adaptively selecting two different kernels to capture homophily and heterophily information. Meanwhile, the learning process of the homophily degree matrix and the bi-kernel feature propagation process are trained jointly to enhance each other in an end-to-end fashion. Extensive experiments on three public data sets demonstrate the effectiveness of the proposed method. Full article

The main goal of the analysis presented in this paper is to examine the dynamics of including other-sex peers in the peer networks of early adolescents, aged 11 (at T1) and 13 (at T2), and the relationship between sex heterophily and changes in the sense of peer integration. The analysis was conducted using the Latent Difference Score (LDS) model with data from a representative nationwide longitudinal study in Poland (n = 5748). With reference to the dynamics related to the heterophilic process, the research confirmed that at the beginning of grade 5 of primary school, heterophily is still relatively rare, yet towards the end of early adolescence, there is a gradual shift, more strongly in girls, towards breaking through the strictly same-sex segregation and embarking on heterophilic relationships. Importantly, the LDS model—even when controlling for different measures of peer network—showed significant and positive (among both girls and boys) relations between establishing cross-sex relationships and the sense of peer integration. The results indicate that the appearance of the opposite sex in the peer network between grades 5 and 6 will improve the sense of peer integration. The findings are discussed in relation to results from other studies in the field. Full article

Social homogeneity, understood as the similarity of perceptions and attitudes that individuals display toward the environment around them, is explained by the relational context in which they are immersed. However, there is no consensus about which relational mechanism best explains social homogeneity. The purpose of this research is to find out which of the three classical relational processes most studied in network analysis (structural cohesion, role equivalence, or homophily) is more determinant in explaining social homogeneity. To achieve the research objective, 110 professionals (psychologists, social workers, and community facilitators) implementing a psychosocial care program in three regions of Northwest Colombia were interviewed. Different types of relationships among professionals were analyzed using network analysis techniques. To examine the structural cohesion hypothesis, interveners were categorized according to the level of structural cohesion by performing core-periphery analysis in the networks evaluated; to test the role equivalence hypothesis, participants were categorized according to their level of degree centrality in the networks examined; to test the homophily hypothesis, participants were grouped according to the level of homophily in terms of professional profile. The non-parametric tests showed that role equivalence was the most powerful mechanism for explaining social homogeneity in the sample of psychosocial interveners evaluated. Full article

Graph convolutional networks (GCNs) have been successfully applied to learning tasks on graph-structured data. However, most traditional GCNs based on graph convolutions assume homophily in graphs, which leads to a poor performance when dealing with heterophilic graphs. Although many novel methods have recently been proposed to deal with heterophily, the effect of homophily and heterophily on classifying node pairs is not clearly separated in existing approaches and inevitably influences each other. To deal with various types of graphs more accurately, in this work we propose a new GCN-based model that leverages the explicitly estimated homophily and heterophily degree between node pairs and adaptively guides the propagation and aggregation of signed messages. We also design a pre-training process to learn homophily and heterophily degree from both original node attributes that are graph-agnostic and the localized graph structure information by using Deepwalk that reflects graph topology. Extensive experiments on eight real-world benchmarks demonstrate that the new approach achieves state-of-the-art results on three homophilic graph datasets and outperforms baselines on five heterophilic graph datasets. Full article

This study explores whether social interaction with dissimilar others can lead to pro-environmental behavior. Dissimilar others are people who differ from the person in question (e.g., in terms of lifestyle or culture). While most research focuses on homogenous social networks (e.g., spatial communities), we explore the potential of network heterophily. Specifically, using data (n = 1370) from the Flemish Survey on Sociocultural Shifts, we examine the relationship between network heterophily and pro-environmental behavior (i.e., shopping decisions and curtailment behavior). Building on Granovetter’s study on ‘the strength of weak ties’, we emphasize the importance of social ties that provide novel information and social expectations. Through interaction with dissimilar others, people may create a heterogeneous network in which a diversity of information and social expectations with regard to pro-environmental behavior circulates. We expect that network heterophily may foster pro-environmental behavior. Our findings indicate that pro-environmental behavior may indeed be positively related to interaction with dissimilar others, partly because people with many dissimilar ties know more about environmental problems and are more concerned about them. This study therefore shows that network heterophily promotes pro-environmental behavior. The paper concludes with a discussion of the functionality of dissimilarity and some avenues for future research. Full article

Inter-organizational networks represent one of the most promising practice-based approaches in public health as a way to attain resources, share knowledge, and, in turn, improve population health outcomes. However, the interdependencies and effectiveness related to the structure, management, and costs of these networks represents a critical item to be addressed. The objective of this research is to identify and determine the extent to which potential partnering patterns influence the structure of collaborative networks. This study examines data collected by PARTNER, specifically public health networks (n = 162), to better understand the structured relationships and interactions among public health organizations and their partners, in relation to collaborative activities. Combined with descriptive analysis, we focus on the composition of public health collaboratives in a series of Exponential Random Graph (ERG) models to examine the partnerships between different organization types to identify the attribute-based effects promoting the formation of network ties within and across collaboratives. We found high variation within and between these collaboratives including composition, diversity, and interactions. The findings of this research suggest common and frequent types of partnerships, as well as opportunities to develop new collaborations. The result of this analysis offer additional evidence to inform and strengthen public health practice partnerships. Full article