A novel method based on GNN and auto-encoder which can search keywords over an imcomplete graph (a graph with missing information).
The datasets except CiteSeer only include the original dataset. Please follow the setting of our work to hide keywords and edges.
The supplementary materials of this paper can be found via the following link: https://proceedings.neurips.cc/paper/2021/file/0d7363894acdee742caf7fe4e97c4d49-Supplemental.pdf
If you use this code in your work, please cite our NeursIPS 2021 paper:
@inproceedings{NEURIPS2021_0d736389,
author = {HAO, YU and Cao, Xin and Sheng, Yufan and Fang, Yixiang and Wang, Wei},
booktitle = {Advances in Neural Information Processing Systems},
editor = {M. Ranzato and A. Beygelzimer and Y. Dauphin and P.S. Liang and J. Wortman Vaughan},
pages = {1700--1712},
publisher = {Curran Associates, Inc.},
title = {KS-GNN: Keywords Search over Incomplete Graphs via Graphs Neural Network},
url = {https://proceedings.neurips.cc/paper/2021/file/0d7363894acdee742caf7fe4e97c4d49-Paper.pdf},
volume = {34},
year = {2021}
}