Thank you for the datasets you have provided, but how to utilise these datasets, does this file just contain the node features?

[CharliedoD]

Thank you for the datasets you have provided, but how to utilise these datasets, does this file just contain the node features?

[EdisonLeeeee]

Hi, @CharliedoD

This repo is a backup of datasets from the paper: Spatio-Temporal Attentive RNN for Node Classification in Temporal Attributed Graphs

You can load the datasets via:

import numpy as np
file = np.load(dataset_path)

x = file['attmats']  # (N, T, D) node features
y = file['labels']  # (N, C)  node labels
adjs = file['adjs']  # (T, N, N) graph snapshots

[CharliedoD]

Thank you very much, I would also like to know if these node features are structure-independent features or are they generated by deepwalk as used by the SpikeNet framework.

[EdisonLeeeee]

The node attributes were generated using word2vec or PCA and are structure-independent. You can kindly refer to the original paper for more details.