benedekrozemberczki / SimGNN

A PyTorch implementation of "SimGNN: A Neural Network Approach to Fast Graph Similarity Computation" (WSDM 2019).
GNU General Public License v3.0
759 stars 147 forks source link

How to make a dataset by myself? #4

Closed TTtaosen closed 5 years ago

TTtaosen commented 5 years ago

hello: i'm trying to using simgnn to run a simple project. but i don't know how to product .json from a graph to it.it's greatful if you can give me some suggestion. thank you.

benedekrozemberczki commented 5 years ago
  1. Create a set of graphs.
  2. For each graph randomly remove and add edges such way that no edge is re-added-re-deleted. The number of such insertions and deletions is the ged.
  3. Add the original and the modified graph to the json as edge lists. Add the GED as the target.