Disentangled Graph Spectral Domain Adaptation
Authors: Liang Yang, Xin Chen, Jiaming Zhuo, Di Jin, Chuan Wang, Xiaochun Cao, Zhen Wang, Yuanfang Guo
ICML 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Quantitative and qualitative experiments justify the superiority of the proposed DGSDA. Section 5 'Experiments' covers experimental setup, result analysis, effectiveness study, ablation study, and hyperparameter analysis, presenting tables and figures with empirical results. |
| Researcher Affiliation | Academia | 1Hebei Province Key Laboratory of Big Data Calculation, School of Artificial Intelligence, Hebei University of Technology, Tianjin, China 2College of Intelligence and Computing, Tianjin University, Tianjin, China 3School of Computer Science and Technology, Beijing Jiao Tong University, Beijing, China 4School of Cyber Science and Technology, Shenzhen Campus of Sun Yatsen University, Shenzhen, China 5School of Artificial Intelligence, OPtics and Electro Nics (i OPEN), School of Cybersecurity, Northwestern Polytechnical University, Xi an, China 6School of Computer Science and Engineering, Beihang University, Beijing, China. Correspondence to: Di Jin <EMAIL>. |
| Pseudocode | No | The paper does not contain a clearly labeled 'Pseudocode' or 'Algorithm' block with structured steps. Section 4.3 'Objective Function and Algorithm' discusses the objective function and its components in text but does not provide a formal algorithm block. |
| Open Source Code | Yes | Our code is available at https://github.com/Hechriver/DGSDA |
| Open Datasets | Yes | The experiment utilizes three types of benchmark datasets: Citation networks (Arnet Miner: ACMv9 (A), Citationv1 (C) and DBLPv7 (D)), social interactions (Blog Catalog and Twitch-DE/EN), and transportation systems (Airport: Brazil (B), Europe (E) and USA (U)). Refer to Appendix C for dataset details. Arnet Miner These datasets are three citation networks obtain from Arnet Miner (Dai et al., 2022). Blog These datasets are derived from the Blog Catalog dataset (Shen et al., 2020). Twitch These datasets are Twitch gamer networks from six regions (Liu et al., 2024a). Airport These datasets are airport traffic networks from three countries (Ribeiro et al., 2017). |
| Dataset Splits | No | The paper discusses the use of labeled source graphs and unlabeled target graphs for Unsupervised Graph Domain Adaptation but does not explicitly provide specific training/test/validation split percentages, counts, or references to predefined splits for the datasets used. For example, it does not specify what percentage of nodes are used for training, validation, or testing within each domain. |
| Hardware Specification | Yes | All experiments are performed on Nvidia Ge Force RTX 3090 (24GB). |
| Software Dependencies | No | Our proposed model DGSDA1 is implemented with Py Torch (Paszke et al., 2017) and Py Torch Geometric library (Fey & Lenssen, 2019). The paper mentions the software packages PyTorch and PyTorch Geometric and cites relevant papers, but does not specify their version numbers. |
| Experiment Setup | Yes | For hyperparameter settings, The node representation dimension is selected from {128, 256}. The learning rate is tuned from {0.01, 0.005, 0.001, 0.0005}, weight decay is tuned from {0.0005, 0.005, 0.01} and K is selected from {5, 8, 10, 15}. |