Graph Structure Learning with Variational Information Bottleneck

Authors: Qingyun Sun, Jianxin Li, Hao Peng, Jia Wu, Xingcheng Fu, Cheng Ji, Philip S Yu4165-4174

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experimental results demonstrate that the superior effectiveness and robustness of VIB-GSL.
Researcher Affiliation Academia 1 Beijing Advanced Innovation Center for Big Data and Brain Computing, Beihang University, Beijing 100191, China 2 School of Computer Science and Engineering, Beihang University, Beijing 100191, China 3 Shenyuan Honors College, Beihang University, Beijing 100191, China 4 School of Computing, Macquarie University, Sydney, Australia 5 Department of Computer Science, University of Illinois at Chicago, Chicago, USA
Pseudocode Yes Algorithm 1: The overall process of VIB-GSL
Open Source Code Yes 1Code is available at https://github.com/VIB-GSL/VIB-GSL.
Open Datasets Yes We empirically perform experiments on VIBGSL on four widely-used social datasets including IMDBB, IMDB-M, REDDIT-B, and COLLAB (Rossi and Ahmed 2015).
Dataset Splits Yes We perform 10-fold cross-validation and report the average accuracy and the standard deviation across the 10 folds in Table 1
Hardware Specification No The paper does not provide specific details on the hardware used for experiments.
Software Dependencies No The paper does not specify version numbers for any software dependencies.
Experiment Setup Yes Parameter Settings. We set both the information bottleneck size K and the embedding dimension of baseline methods as 16. For VIB-GSL, we set t = 0.1 in Eq. (13), a0 = 0.1 and perform hyperparameter search of β {10 1, 10 2, 10 3, 10 4, 10 5, 10 6} for each dataset.