aiqbench analysis
This research addresses the critical challenge of efficient "unlearning" in spatio-temporal graph models, enabling compliance with privacy regulations like GDPR without requiring costly full model retraining. By introducing a biologically inspired framework called CallosumNet, the authors demonstrate how to erase specific node data while preserving global accuracy across dynamic processes such as healthcare monitoring and molecular dynamics. The approach offers a significant advancement in balancing data utility with strict privacy constraints within complex graph neural networks.
Unlearning on Spatio-Temporal Graphs through Subgraph Virtual Edge Reconstruction
arXiv:2608.29369v1 Announce Type: new Abstract: Spatio-temporal graphs are widely used in modeling complex dynamic processes such as temporal forecasting, molecular dynamics, and healthcare monitoring. Recently, stringent privacy regulations such as GDPR and CCPA have introduced significant new challenges for existing spatio-temporal graph models, requiring complete unlearning of unauthorized data. Since each node in a spatio-temporal graph diffuses information globally across both spatial and temporal dimensions, existing unlearning methods primarily designed for static graphs and localized data removal cannot efficiently erase a single node without incurring costs nearly equivalent to full model retraining. To address this, we propose CallosumNet, a spatio-temporal graph unlearning framework biologically inspired by the corpus callosum structure. CallosumNet makes two key technical contributions: (1) it reconstructs subgraphs using biologically-inspired virtual edges; and (2) it restores interlinked spatio-temporal dependencies among subgraphs via a lightweight meta-graph integration layer. Empirical results on four diverse real-world datasets show that CallosumNet achieves complete unlearning while maintaining accuracy very close to the gold model. The code is publicly available at https://github.com/wenlu-lab/STGraphUnlearning.
Abstract:Spatio-temporal graphs are widely used in modeling complex dynamic processes such as temporal forecasting, molecular dynamics, and healthcare monitoring. Recently, stringent privacy regulations such as GDPR and CCPA have introduced significant new challenges for existing spatio-temporal graph models, requiring complete unlearning of unauthorized data. Since each node in a spatio-temporal graph diffuses information globally across both spatial and temporal dimensions, existing unlearning methods primarily designed for static graphs and localized data removal cannot efficiently erase a single node without incurring costs nearly equivalent to full model retraining. To address this, we propose CallosumNet, a spatio-temporal graph unlearning framework biologically inspired by the corpus callosum structure. CallosumNet makes two key technical contributions: (1) it reconstructs subgraphs using biologically-inspired virtual edges; and (2) it restores interlinked spatio-temporal dependencies among subgraphs via a lightweight meta-graph integration layer. Empirical results on four diverse real-world datasets show that CallosumNet achieves complete unlearning while maintaining accuracy very close to the gold model. The code is publicly available at this https URL.
Accepted as a short paper at ACM SIGSPATIAL 2026. 4 pages
Machine Learning (cs.LG); Cryptography and Security (cs.CR)
arXiv:2608.29369 [cs.LG]
(or arXiv:2608.29369v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2608.29369 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 274 words
Related coverage
Spatial Entropy based Partitioning for Spatiotemporal Graph Unlearning
arXiv · Sep 1, 2026
Graph-centric agentic intelligence
Amazon Science · Oct 1, 2026
PPO-STGNN: A Proximal Policy Optimization Approach with Spatio-Temporal Graph Neural Networks for DAG Task Scheduling in Cloud-Edge-End Computing
arXiv · Sep 5, 2026
Transfer Learning for Edge Classification on Dynamic Text-Attributed Graphs
arXiv · Sep 30, 2026