djasingh / resip_dataset Goto Github PK

Instruction:

1. Unzip dataset.tar.bz2
2. The folder contains multiple numpy or pickle files
3. The file names show the dataset used with methods described in the paper.
4. For ResIP-M1 and ResIP-M2 the train and test data contains input, single output (ResIP-M1), and output sequence (ResIP-M2) in tuple
5. For ResIP-M3, the file names contains "ResIP-M3".
6. The output is not scaled (use log2 transformation)
7. The input represents sequence of embedding vectors for the researchers in the family graph/tree
8. The output values represent family size during prediction period.
9. For detail check the ResIP paper: https://doi.org/10.1016/j.joi.2023.101392