A collection of temporal graph datasets from Towards Better Evaluation for Dynamic Link Prediction with pytorch-geometric TemporalData class and TemporalDataloader.
BUT, this repository is not official, some of the datasets are not included, and the preprocessing is not the same as the original paper.
pip install git+https://github.com/hsiangjenli/temporal-graph-dataset.git
from temporal_graph import TemporalGraphDataset
tgd = TemporalGraphDataset(root="data")tgd.available_datasetsdata, x, train_mask, val_mask, test_mask = tgd("tgbl-flight")- torch_geometric.loader.TemporalDataLoader
from torch_geometric.loader import TemporalDataLoader
loader = TemporalDataLoader(data[train_mask], batch_size=1000, neg_sampling_ratio=1.0)temporal_graph/preprocess/p_{dataset_name}.pyis a Python script to preprocess the dataset. To see the details, please check the Makefile.
- Dataset for "Towards Better Evaluation for Dynamic Link Prediction"
- https://zenodo.org/records/7213796#.Y1cO6y8r30o
- Poursafaei, Farimah and Huang, Shenyang and Pelrine, Kellin and and Rabbany, Reihaneh
This project is part of my master's thesis. Feel free to use or modify the code, but please keep in mind that I'm just a student, so there might be mistakes. If you notice any bugs or have suggestions for improvement, please feel free to open an issue or pull request. Thanks for your understanding and support!"
RN Lee @ 2024-05 😊