Adding a new dataset

[InesBlin]

Hi, thanks a lot for your interesting work.

I'm looking at running your method on custom datasets for a semi-inductive setting, what would be your recommendation? At the moment, it looks like the datasets are tied to google drive files.

Also, what's the difference in your case between triples and statements?

Is it ok to provide the statements file only (with files transductive_train, inductive_train, inductive_ts, inductive_val) , and then your method will add the index and embeddings in the background?

Thanks a lot for your help,
Best,
Inès