Download datasets and put them into the /Dataset folder.
Each multilayer graph should be stored in a .txt file, and the filename is considered the dataset name.
The first line of the .txt file should contain #layers, #vertex, #edges.
Following the header, each row contains three integers, which are layer-id, node-id, node-id.
An example file "homo" is provided, and new datasets should adhere to the same structure.
We also include a script to convert WikiTalk and StackOverflow from original snap format (Wiki, SO) to multilayer graph. Refer to "ConvertWiki.py" and "CovertSO.py" for usage.
The two case study datasets are also included as dblp-ijcai-kdd-mod.txt and dblp-ijcai-kdd-mod-aaai.txt. metainfo_id_author.txt shows a map from node-id to author name for dblp-ijcai-kdd-mod.txt. For more information about how to convert DBLP dataset to multilayer graphs, check our previous code at here.
The missing datasets are available at FirmCore and multilayer kCore.
Our code provides an acceleration using Cython as follows:
python setup.py build_ext --inplaceWithout using Cython, the codes also works.
And you can use "clean.py" to clean the files created by Cython.
In general, the code can be run as follows:
python main.py -a [Decomposition/Subgraph] -d [dataset name(default homo)] -m [EP/IP/VC(default)]python main.py -a Decomposition -d [dataset(default homo)] -m [EP/IP/VC(default)]examples:
Using Vertex Centric to compute FoCore Decomposition of Homo Dataset:
python main.py -a Decomposition -d homo -m VCUsing Interleaved Peeling to compute FoCore Decomposition of Amazon Dataset:
python main.py -a Decomposition -d Amazon -m IPpython main.py -a Subgraph -d [dataset(default homo)]examples:
Compute denest subgraph of Homo Dataset:
python main.py -a Subgraph -d homo@inproceedings{wangFocusCoreDecompositionMultilayer2024,
title = {{{FocusCore Decomposition}} of {{Multilayer Graphs}}},
booktitle = {2024 {{IEEE}} 40th {{International Conference}} on {{Data Engineering}} ({{ICDE}})},
author = {Wang, Run-An and Liu, Dandan and Zou, Zhaonian},
year = {2024},
pages = {2792--2804},
issn = {2375-026X},
doi = {10.1109/ICDE60146.2024.00218},
url = {https://ieeexplore.ieee.org/document/10597766},
}