search

LuckyTiger123 / GraphTCM

The code Implementation of the paper “Exploring Correlations of Self-Supervised Tasks for Graphs”.
4 stars 1 forks source link

The dataset split for link prediction #1

Closed lei-liu1 closed 6 months ago

lei-liu1 commented 6 months ago

Hello,

Thanks for sharing the codes. I would like to know the dataset split ratio for the link prediction task, but I couldn't find it in the paper or the codes. Could you please share the split ratio of the dataset used for the results reported in the paper?

Best Regards, Lei

LuckyTiger123 commented 6 months ago

Thanks for your interest in our work. In link prediction tasks on Cora, CiteSeer, and PubMed, we randomly selected 5% as the validation set and 10% as the test set. We conducted link prediction experiments following the procedure outlined in https://github.com/pyg-team/pytorch_geometric/blob/master/examples/link_pred.py. We hope our response can address your issue.

lei-liu1 commented 6 months ago

Thanks!