DRGW: Learning Disentangled Representations for Robust Graph Watermarking

2026-01-20

AI Summary

Existing graph watermarking methods struggle to simultaneously achieve transparency, robustness, and controllable discrete structural modifications due to their direct manipulation of graph topology or entangled representations. This work proposes the first decoupled representation learning–based graph watermarking framework (DRGW), which employs adversarial training to disentangle invariant structural features from an independent watermark carrier. By integrating a graph-aware invertible neural network, DRGW enables lossless watermark embedding and extraction, while a structure-aware editor precisely maps latent-space perturbations to discrete graph edits. The method achieves statistical disentanglement between watermark and structural information for the first time, significantly outperforming state-of-the-art approaches across multiple benchmark datasets and demonstrating breakthrough performance in robustness, transparency, and detectability.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Computer Vision: Adversarial Attacks & Robustness
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Web Mining and Content Analysis: Bridging structured and unstructured data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Graph-structured data is foundational to numerous web applications, and watermarking is crucial for protecting their intellectual property and ensuring data provenance. Existing watermarking methods primarily operate on graph structures or entangled graph representations, which compromise the transparency and robustness of watermarks due to the information coupling in representing graphs and uncontrollable discretization in transforming continuous numerical representations into graph structures. This motivates us to propose DRGW, the first graph watermarking framework that addresses these issues through disentangled representation learning. Specifically, we design an adversarially trained encoder that learns an invariant structural representation against diverse perturbations and derives a statistically independent watermark carrier, ensuring both robustness and transparency of watermarks. Meanwhile, we devise a graph-aware invertible neural network to provide a lossless channel for watermark embedding and extraction, guaranteeing high detectability and transparency of watermarks. Additionally, we develop a structure-aware editor that resolves the issue of latent modifications into discrete graph edits, ensuring robustness against structural perturbations. Experiments on diverse benchmark datasets demonstrate the superior effectiveness of DRGW.

Problem

• graph watermarking
• disentangled representation
• robustness
• transparency
• structural perturbations

Innovation

• Disentangled Representation
• Graph Watermarking
• Invertible Neural Network
• Adversarial Training
• Structure-aware Editing