pprei's Introduction

Towards Deeper Understanding of PPR-based Embedding Approaches: A Topological Perspective (PPREI)

This is our Pytorch implementation for the paper:

Xingyi Zhang, Zixuan Weng, Sibo Wang. Towards Deeper Understanding of PPR-based Embedding Approaches: A Topological Perspective." WWW 2024.

Requirements:

Python=3.10.12
PyTorch=1.12.0
Scipy=1.10.1
networkx=2.8.4
numpy=1.24.3

Preprocessing the dataset:

run processData

Parameters

f: the name of dataset
r: the rank of matrix

python processData.py -f Brazil

Training:

run PPR

Parameters

f: the name of dataset
r: the rank of matrix
a: the value of teleport probability
i: the number of training epoch
t: the step of PPR
e: the value of threshould

python  PPR.py -f Brazil.mat -r 128 -a 0.7 -i 40 -t 10 -e 1e-7

run network_stats_PPR

Parameters

f: the name of dataset
r: the rank of matrix

python network_stats_PPR.py -f Brazil.mat -r 128

Download datasets

https://drive.google.com/drive/folders/1zsCFA7U8ZKV1bg5g9IwNxqf1TzgGwokw?usp=sharing

Acknowledgements:

Part of the code refers to https://github.com/konsotirop/Invert_Embeddings, which is the official implementation of the paper:

Sudhanshu Chanpuriya, Cameron Musco, Konstantinos Sotiropoulos, Charalampos E. Tsourakakis. "DeepWalking Backwards: From Embeddings Back to Graphs." ICML 2021.

Reference:

Any scientific publications that use our codes should cite the following paper as the reference:

 @inproceedings{zhang&weng2024PPREI,
     title = "Towards Deeper Understanding of PPR-based Embedding Approaches: A Topological Perspective",
     author = {Xingyi Zhang and
     		  Zixuan Weng and
     		  Sibo wang},
     booktitle = {{WWW}},
     year = {2024},
 }

If you have any questions for our paper or codes, please send an email to [email protected].

Recommend Projects