This repo is for reproducing our work High-quality Task Division for Large-scale Entity Alignment, which has been accepted at CIKM 2022 (arXiv).
Download the code and data. The structure of folders should be organized as below
divea/
|- datasets/ # datasets are put under this folder
|- dbp15k/
|- dwy100k/
|- 2m/ # dataset fb_dbp of size 2M
|- divea/ # code of our method
|- RREA/ # RREA model
|- GCN-Align/ # GCN-Align model
|- scripts/ # scripts files for running our method with RREA
|- scripts2/ # scripts files for running our method with GCN-Align
|- environment.yml # conda environment file
|- README.md
cd to project directory firstly.
Create the environment named divea and install most packages by running command:
conda env create -f environment.ymlThen, activate the environment:
conda activate diveaFinally, install package networkx-metis as below. Other installation instructions of networkx-metis can be found here.
git clone https://github.com/networkx/networkx-metis.git
cd networkx-metis/
python setup.py build
python setup.py installThe scripts for running our method with RREA are put under scripts/.
bash run_over_perf_vs_cps.sh. Overall performance. Table 1.bash run_over_perf_vs_sbp.sh. Overall performance. Table 2.bash run_over_perf_vs_cps_2m.sh. Overall performance. Table 1.bash run_over_perf_vs_sbp_2m.sh. Overall performance. Table 2.
The scripts for running our method with GCN-Align are put under scripts2/. The script file names and corresponding functions can be aligned with scripts under scripts/.
Please cite this paper if you use the released code in your work.
@inproceedings{DBLP:conf/cikm/LiuHZZZ22,
author = {Bing Liu and
Wen Hua and
Guido Zuccon and
Genghong Zhao and
Xia Zhang},
editor = {Mohammad Al Hasan and
Li Xiong},
title = {High-quality Task Division for Large-scale Entity Alignment},
booktitle = {Proceedings of the 31st {ACM} International Conference on Information
{\&} Knowledge Management, Atlanta, GA, USA, October 17-21, 2022},
pages = {1258--1268},
publisher = {{ACM}},
year = {2022},
url = {https://doi.org/10.1145/3511808.3557352},
doi = {10.1145/3511808.3557352},
timestamp = {Wed, 04 Jan 2023 07:33:22 +0100},
biburl = {https://dblp.org/rec/conf/cikm/LiuHZZZ22.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent