This survey presents the first systematic review of DHG representation learning methods and proposes a novel algorithm-centric taxonomy that categorizes existing literature, including early embedding-based approaches, graph neural network (GNN)-based models, and relatively recent Transformer-based DHG methods, while explicitly highlighting their intrinsic modeling biases with respect to dynamic granularity.
Abstract
Graph representation learning (GRL) serves as a canonical paradigm for modeling complex networks. However, real-world AI systems inherently manifest as evolving heterogeneous entities with complex interactions, posing significant challenges to static or homogeneous modeling. To address these complexities, representation learning for Dynamic Heterogeneous Graphs (DHGs) has emerged as a vital approach for learning low-dimensional representations that simultaneously preserve structural semantics and temporal dynamics. This survey presents the first systematic review of DHG representation learning methods. We first introduce a unified formal definition that encompasses both discrete-time and continuous-time DHGs from the perspective of temporal granularity. Building upon this formulation, we propose a novel algorithm-centric taxonomy that categorizes existing literature, including early embedding-based approaches, graph neural network (GNN)-based models, and relatively recent Transformer-based DHG methods, while explicitly highlighting their intrinsic modeling biases with respect to dynamic granularity. Furthermore, we summarize representative applications of DHG representation learning, along with commonly used datasets and benchmarks. Finally, we discuss promising research directions that guide future advances in this rapidly evolving field.
Dynamic heterogeneous graphs, as an important tool for characterizing multiple types of entities and their evolutionary relationships in complex systems, have been widely used in scenarios such as academic network analysis, e-commerce, and social networks. However, compared to static homogeneous graphs, the heterogeneity and temporal dynamics in dynamic heterogeneous graph data pose additional challenges for representation learning, meaning that existing dynamic heterogeneous graph learning methods still encounter the following challenges during modeling: (1) There is a problem of limited semantic propagation of meta-paths within time snapshots; (2) There is a low-order relationship bias problem caused by dynamic optimization objectives. To address the above challenges, we propose a dynamic heterogeneous graph neural network based on multi-view contrastive learning (DHGNN-MVC). Specifically, we explicitly introduce the attribute information of intermediate nodes in the meta-path during the heterogeneous information aggregation process by optimizing the meta-path semantic modeling approach. Combined with a message propagation mechanism based on the PPR matrix, this allows us to collaboratively capture both local and deep information in heterogeneous graph snapshots from the perspectives of relationships and meta-paths, thereby effectively alleviating the limitations of meta-path semantic transmission within temporal snapshots. At the same time, we propose a multi-view contrastive learning strategy that jointly optimizes the future graph reconstruction loss, local-global contrastive loss, and community evolution contrastive loss. The node representations are constrained at three scales: local structure, graph level semantics, and community level to effectively alleviate the problem of low-order relationship bias. The simulation experiments conducted on multiple real-world dynamic heterogeneous graph datasets show that the performance of DHGNN-MVC is significantly better than existing baseline methods, fully demonstrating its effectiveness and novelty in capturing the evolution laws of dynamic heterogeneous graph networks.
Graph Autoencoder Uncovering Descriptive Information (GAUDI), an unsupervised graph deep learning framework designed to capture both local details and global structure, is introduced, providing new insights into emergent phenomena across diverse scientific domains.
Heterogeneous graphs are a widely adopted data representation paradigm owing to their ability to model multiple node types and complex relationships uniformly. However, due to privacy and ownership restrictions, semantic-rich heterogeneous graphs are often stored by multiple participants, each of whom holds a subgraph. Federated heterogeneous subgraph learning seeks to aggregate knowledge from structurally diverse, decentralized subgraphs. This collaborative approach trains a more robust graph model without exposing raw data, thus overcoming the challenges of modeling complex relationships. In practical applications, however, the incompleteness and heterogeneity of subgraphs often result in degraded performance of the federated heterogeneous subgraph learning model. The method based on structural complementarity for addressing structural incompleteness ignores the semantic rules and node roles that are unique to heterogeneous graphs. For heterogeneous subgraphs, the information compression method based on category labels loses key high-order topological semantics. Therefore, we propose a new federated learning framework, federated heterogeneous subgraph learning with meta-path-guided role mapping and condensation (FedPRC). Specifically, based on the principle of structural equivalence, we propose a meta-path-guided virtual node generation method. This method enables each client to effectively infer and complete missing semantic links locally. Based on the evaluation of the importance of subgraph structure, we design a meta-path-guided information-aware subgraph condensation method. This method uses the condensation subgraph as an effective carrier for cross-client knowledge alignment and communication, while retaining the key topology and high-order paths. Our experiments on multiple real-world graph datasets demonstrate that this framework significantly outperforms existing federated graph learning baseline models in node classification tasks, validating its effectiveness.
Yanjin Cheng, Wenmin Li, Su-Lu Qin et al.· Journal of King Saud Univers...· 0 citations
This work introduces Node4All, a node representation learner applicable to arbitrary graph datasets without any dataset-specific optimization, and introduces the Channel Graph Transformer (CGT), which enables a single fixed parameterization to process arbitrary graph datasets.
Dooho Lee, Jaemin Yoo· Proceedings of the 32nd ACM...· 0 citations
Learning graph structures from data is a fundamental problem that spans a wide range of signal processing and machine learning tasks. While significant effort has been made to tackle the problem, existing research has largely evolved along two parallel directions. The first seeks to infer the topology of an individual graph from observations supported on it, whereas the second seeks to learn a generative distribution from observed graph instances, enabling the sampling of new graphs. This review presents a unified framework that connects these formulations by viewing them as inverse problems of a common generation process for graph data. We review the major methodologies within this framework, highlight their relationships, strengths, and limitations, and identify opportunities for integrating ideas across paradigms. By bridging graph topology learning and graph generation, this review provides a broader cross-disciplinary perspective on the field and outlines promising directions for future research.
Xiaowen Dong, Hoi-To Wai, Si-Heng Chen et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
