/UnderstandingNetworkEmbedding

"On Interpretation of Network Embedding via Taxonomy Induction"

Primary LanguageMatlab

UnderstandingNetworkEmbedding

This project implements the proposed framework for running the experiments in the paper:

On Interpretation of Network Embedding via Taxonomy Induction
Ninghao Liu, Xiao Huang, Jundong Li, Xia Hu
Proceedings of KDD'18, London, UK

Please cite our paper if you use the codes. Thanks!

Example to run the codes in MATLAB:

[Beta_est, G, tree, splits, is_leaf, clusters, priorities, Y] = UnderstandingNetworkEmbedding(Emb, X, C, name_net, name_embd)

Some input variable and parameters are introduced as below:

  • Emb: The embedding matrix (D dimension * N data points)
  • X: The attribute matrix (N data points * M attributes)
  • C: The number of leaf nodes to be generated in the taxonomy
  • name_net: A string about the name of the dataset
  • name_embd: A string about the name of the embedding model