GraphProt: Certified Black-Box Shielding Against Backdoored Graph Models

Authors: Xiao Yang, Yuni Lai, Kai Zhou, Gaolei Li, Jianhua Li, Hang Zhang

IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we present the results of our comparative analyses and ablation studies on GRAPHPROT. Notably, it functions within rigorous black-box conditions (with access limited to the test graph and few queries). Consequently, we primarily assess whether our approach achieves comparable efficacy to current defense strategies. ... Empirical evaluations conducted on multiple attack paradigms and benchmark datasets confirm its effectiveness in reducing attack success rates while preserving benign data accuracy.
Researcher Affiliation Academia 1Shanghai Jiao Tong University 2Hong Kong Polytechnic University 3Cornell University {youngshall, gaolei li, lijh888}sjtu.edu.cn, EMAIL, EMAIL, EMAIL
Pseudocode No The paper describes the methodology in text and uses a figure (Figure 1) to depict the workflow, but no structured pseudocode or algorithm blocks are provided.
Open Source Code No The paper does not contain any explicit statements about releasing source code or provide links to a code repository.
Open Datasets Yes We employed 6 benchmark datasets: AIDS [Rossi and Ahmed, 2015], ENZYMES [Dobson and Doig, 2003], DHFR [Morris et al., 2020], NCI1 [Wale and Karypis, 2006], PROTEINS [Borgwardt et al., 2005], and COLLAB [Yanardag and Vishwanathan, 2015].
Dataset Splits Yes For each dataset, we randomly allocated two-thirds of the graphs for training the backdoored victim model, preserving the remainder for empirical testing.
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments (e.g., GPU models, CPU types).
Software Dependencies No The paper mentions GNN models like GCN, SAGE, and GAT, but does not specify any software libraries or frameworks with their version numbers.
Experiment Setup Yes The GRAPHPROT configuration includes the subgraph number K= 5, the samplerate p = 0.2, and the feature selection proportion r = 0.8. ... Multiple subgraph numbers K were configured at intervals of 3, ranging from 1 to 22... Sample-rates (0 – 100%) were applied for GRAPHPROT-R... We subsequently adjusted the feature fraction r of GRAPHPROT-TF across [0, 100]%...