STEA

Code for COLING2022 paper “A Simple Temporal Information Matching Mechanism for Entity Alignment Between Temporal Knowledge Graphs”.

Datasets

The datasets are from TEA-GNN.
ent_ids_1: ids for entities in source KG;
ent_ids_2: ids for entities in target KG;
triples_1: relation triples encoded by ids in source KG;
triples_2: relation triples encoded by ids in target KG;
rel_ids_1: ids for entities in source KG;
rel_ids_2: ids for entities in target KG;
sup_pairs + ref_pairs: entity alignments

Environment

Anaconda>=4.5.11
Python>=3.7.11
pytorch>=1.10.1

Usage

Use the following command:
python main.py
Before conducting unsupervised experiments, you need to run "get_unsup_seeds.py" to obtain unsupervised seeds.
The obtained unsupervised seed file--"*.npy" is stored in the "simt" folder (the folder need to be added manually) in the corresponding dataset folder.
You also need to set the parameter "unsupervise" to "True" in the "args.py".

Acknowledgement

We refer to the code of RREA. Thanks for their great contributions!

Citation

If you use this model or code, please cite it as follows:
@inproceedings{Cai2022STEA,
author = {Li Cai and Xin Mao and Meirong Ma and Hao Yuan and Jianchao Zhu and Man Lan},
title = {A Simple Temporal Information Matching Mechanism for Entity Alignment Between Temporal Knowledge Graphs},
booktitle = {Proceedings of the 29th International Conference on Computational Linguistics},
pages = {2075--2086},
year = {2022}, }