Skip to content
master
Switch branches/tags
Go to file
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 

README.md

Deep Multi-attributed Graph Translation with Node-Edge Co-evolution.

image_text This repository is the official Tensorflow implementation of NEC-DGT, a multi-attributed graph translation model.

The relevant paper is "Deep Multi-attributed Graph Translation with Node-Edge Co-evolution".

Xiaojie Guo, Liang Zhao, Cameron Nowzari, Setareh Rafatirad, Houman Homayoun, and Sai Dinakarrao (ICDM 2019 Best Paper Award).

Installation

Install Tensorflow following the instuctions on the official website. The code has been tested over Tensorflow 1.13.1 version.

Datasets

The dataset can be found and downloaded at Datasets for Deep Graph Translation. The datasets used here are IoT datasets, B-A datasets, and E-R datasets.

For directly using datasets to run the code without modifying the format, please find datasets here.

Code Description

There are four experiment tasks: synthetic1, synthetic2, Molecule (reactaction prediction) and IoT malware confinement. For each experiment task: main.py is the main executable file which includes specific arguments and training iterations and calls model.py and utlis.py. utlis.py is where the dataset is read.

Run the code

For each task, to train a model,edit the FLAGS "--type" to "train" and:

           python main.py

to test a model,edit the FLAGS "--type" to "test" and:

           python main.py             

Outputs

There are several different outputs to store the generated edges and nodes.

O_t.npy contains the nodes attributes of the generated target graphs.

O_x.npy contains the nodes attributes of the input graphs.

O_y.npy contains the nodes attributes of the real target graphs.

Ra_t.npy contains the edge attributes of the generated target graphs.

Ra_x.npy contains the edge attributes of the input graphs.

Ra_y.npy contains the edge attributes of the real target graphs.

Evaluation

The evaluation is atomatically done and print out at the end of the test process.

About

This is the implementation and supplemental materials for paper "Deep Multi-attributed Graph Translation with Node-Edge Co-evolution"

Resources

Releases

No releases published

Packages

No packages published

Languages