---
title: "ADC-GNN: Adaptive Dual-level Collaborative Graph Neural Networks for Graph Classification"
authors:
  - "Wan Tang"
  - "Lu Bai"
  - "Lixin Cui"
  - "Ming Li"
  - "Hangyuan Du"
  - "Jing Li"
doi: "10.24963/ijcai.2026/337"
url: "https://ijcai.org/proceedings/2026/337"
pdf: "https://ijcai.org/proceedings/2026/0337.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "3033-3042"
track: "Main Track"
keywords:
  - "Data Mining: Mining graphs"
---

# ADC-GNN: Adaptive Dual-level Collaborative Graph Neural Networks for Graph Classification

Wan Tang, Lu Bai, Lixin Cui, Ming Li, Hangyuan Du, Jing Li

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 3033-3042. DOI: [10.24963/ijcai.2026/337](https://doi.org/10.24963/ijcai.2026/337)

## Abstract

Most existing Graph Neural Networks (GNNs) rely on the node-level message passing or attention mechanisms to propagate and extract useful information. Although recent advances attempt to move beyond purely the node-level propagation by constructing high-level representations, these approaches are often constrained by pre-computed substructures or unidirectional bottom-up aggregations. Consequently, high-level structural semantics cannot effectively feed back to guide node representation learning, limiting the collaborative optimization between fine-grained features and macroscopic structural semantics. To address these limitations, we propose a novel Adaptive Dual-level Collaborative GNN (ADC-GNN) associated with an adaptive dual-level collaborative mechanism. We commence by introducing a set of global, learnable latent prototypes as high-level semantic references, and then employ a relaxed Sinkhorn algorithm to establish differentiable, non-collapsing assignments between nodes and prototypes. Based on these assignments, the ADC-GNN constructs high-level representations and enables interactions among them. We show that the ADC-GNN can inject the learned high-level information back into the node level, forming a closed-loop, bidirectional optimization process. Experiments demonstrate the superior performance of the proposed ADC-GNN on graph classification.