Dual Label-Guided Graph Refinement for Multi-View Graph Clustering
Authors: Yawen Ling, Jianpeng Chen, Yazhou Ren, Xiaorong Pu, Jie Xu, Xiaofeng Zhu, Lifang He
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The experimental results show the superior performance on coping with low homophilous graph data. |
| Researcher Affiliation | Academia | 1School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China 2Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China, Shenzhen, China 3Department of Computer Science and Engineering, Lehigh University, Bethlehem, PA, USA |
| Pseudocode | No | The paper describes the model and its components using textual descriptions and mathematical equations, but does not present a formal pseudocode or algorithm block. |
| Open Source Code | Yes | The source code for Dua LGR is available at https://github.com/Yw L-zhufeng/Dua LGR. |
| Open Datasets | Yes | The datasets consist of three categaries i.e., two raw high homophilous datasets, two raw low homophilous datasets and six synthetic datasets. Homophilous graph datasets: two widely used homophilous multi-graph data, including ACM and DBLP. ACM is a paper network from the ACM database1... DBLP is an author network from DBLP database2... Low homophilous graph datasets: Texas and Chameleon are adopted to test the performance of Dua LGR in low homophilous graphs. Texas is a webpage graph from Web KB3... Chameleon (HR 0.23) is a subset of the Wikipedia network (Rozemberczki, Allen, and Sarkar 2021). Footnotes: 1https://dl.acm.org/, 2https://dblp.uni-trier.de/, 3http://www.cs.cmu.edu/afs/cs.cmu.edu/project/theo11/www/wwkb |
| Dataset Splits | No | The paper uses well-known datasets but does not explicitly specify train/validation/test splits, percentages, or sample counts, nor does it refer to standard splits for these datasets. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware (e.g., GPU/CPU models, memory, or specialized accelerators) used to conduct the experiments. |
| Software Dependencies | No | The paper does not list specific software dependencies (e.g., libraries, frameworks) with their version numbers that would be necessary to replicate the experiments. |
| Experiment Setup | Yes | where α is a hyperparameter that controls the influence of the homophily of the graph acorss different views. ... where ε is a cut-off to filter out low homophilous graphs. ... where p is a smooth-sharp parameter. ... Finally, the loss function is formulated as: L = LRec + γLKL, (16) where, γ denotes the trade-off parameter. ... The right side of Figure 2 shows the parameter sensitive analysis about α. α controls the influence of given graphs, which means the larger the α is, the lower the refining extent is, and vice versa. From the right side of Figure 2, on ACM dataset, the highest ACC is 92.1% when α = 1, compared to 91.6% and 91.4% when α = 0.1 and α = 10 respectively. |