Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces CA-DGCL, a novel framework for dynamic graph continual learning that effectively mitigates catastrophic forgetting by leveraging temporal information across graph snapshots. By condensing historical graphs into compact semantic representations and employing Tucker decomposition on a constructed third-order tensor, CA-DGCL stabilizes node features that encapsulate historical knowledge. Extensive experiments reveal that CA-DGCL significantly outperforms existing methods in forgetting suppression while maintaining competitive accuracy, highlighting its potential for dynamic graph applications.
CA-DGCL achieves superior forgetting suppression in dynamic graphs by intelligently condensing historical knowledge without sacrificing performance on new data.
Dynamic graph continual learning (DGCL) is an effective manner for handling catastrophic forgetting in dynamic graphs. However, existing DGCL methods underutilize temporal information across graph snapshots. To address this critical issue, we propose a novel framework for Dynamic Graph Continual Learning via Condensation and Attachment (CA-DGCL). Specifically, CA-DGCL first condenses historical graph snapshots into compact semantic representations efficiently. Further, a cross-timestamp node chains is built to construct a third-order tensor and Tucker decomposition is applied to this tensor for obtaining stable node features, which encapsulate historical knowledge. Finally, these node features are used to generate new nodes and attached to the current graph for replaying of past information without compromising the new patterns. In addtion, a refined forgetting measure is introduced to make it more suitable for dynamic graph settings. Extensive experiments demonstrate that CA-DGCL outperforms baselines in forgetting suppression as well as maintain competitive accuracy, proving its efficacy for dynamic graph continual learning.