search

THUDM / GATNE

Source code and dataset for KDD 2019 paper "Representation Learning for Attributed Multiplex Heterogeneous Network"
MIT License
525 stars 141 forks source link

node embedding #68

Closed kanhuimin closed 4 years ago

kanhuimin commented 4 years ago

Hello, your work is very good. I would like to ask if your work is applicable to the situation that there is only one type of node, and there may be multiple relationships between nodes, but only one kind of relationship association is selected between node pairs? What's more, in your work, you get the embedding of each node for each relationship. Do you support the node embedding of the target node in multiple relationships? We are looking forward to your reply

cenyk1230 commented 4 years ago

Hi @kanhuimin,

Thank you for your interest in our work.

The situation that you say is a little different from our setting. I just come up with a solution to your situation. You can add a base relationship in your graph (two nodes have a base relationship if and only if there is another relationship between them). Then you can use GATNE to generate node embeddings on this extended graph. The node embedding under the base relationship can be used for the downstream node classification task. I think you can have a try if you want to use GATNE models.