Our proposed method REA (Robust Entity Alignment) consists of two components: noise detection and noise-aware entity alignment.

The noise detection is designed by following the adversarial training principle. The noise-aware entity alignment is devised by leveraging graph neural network based knowledge graph encoder as the core. In order to mutually boost the performance of the two components, we propose a unified reinforced training strategy to combine them.

REA is a plug-and-play strategy to mitigate the effect of noise in the given labeled entity pairs for entity alignment problem. The idea also can be easily developed for other alignment algorithms.

Contact: Shichao Pei ([email protected])

• python>=3.5
• tensorflow>=1.10.1
• scipy>=1.1.0
• networkx>=2.2

python3 train.py --lang zh_en

Datasets are from JAPE.

Please refer to our paper.

@inproceedings{pei2020rea,
  title={Rea: Robust cross-lingual entity alignment between knowledge graphs},
  author={Pei, Shichao and Yu, Lu and Yu, Guoxian and Zhang, Xiangliang},
  booktitle={Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery \& Data Mining},
  pages={2175--2184},
  year={2020}
}