The Implementation of "Deep Recursive Network Embedding with Regular Equivalence"(KDD 2018)
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


The Implementation of "Deep Recursive Network Embedding with Regular Equivalence"(KDD 2018).


Python >= 3.5.2
scipy >= 0.19.1
numpy >= 1.13.1
tensorflow == 1.2.0
networkx >= 1.11


Example Usage
python src/ --data_path dataset/barbell.edgelist --save_path result/barbell --save_suffix test \
      -s 16 -b 256 -lr 0.0025 --index_from_0 True
Full Command List
usage: Deep Recursive Network Embedding with Regular Equivalence
       [-h] [--data_path DATA_PATH] [--save_path SAVE_PATH]
       [--save_suffix SAVE_SUFFIX] [-s EMBEDDING_SIZE] [-e EPOCHS_TO_TRAIN]
       [-b BATCH_SIZE] [-lr LEARNING_RATE] [--undirected UNDIRECTED]
       [-a ALPHA] [-l LAMB] [-g GRAD_CLIP] [-K K]
       [--sampling_size SAMPLING_SIZE] [--seed SEED]
       [--index_from_0 INDEX_FROM_0]

optional arguments:
  -h, --help            show this help message and exit
  --data_path DATA_PATH
                        Directory to load data.
  --save_path SAVE_PATH
                        Directory to save data.
  --save_suffix SAVE_SUFFIX
                        Directory to save data.
  -s EMBEDDING_SIZE, --embedding_size EMBEDDING_SIZE
                        the embedding dimension size
  -e EPOCHS_TO_TRAIN, --epochs_to_train EPOCHS_TO_TRAIN
                        Number of epoch to train. Each epoch processes the
                        training data once completely
  -b BATCH_SIZE, --batch_size BATCH_SIZE
                        Number of training examples processed per step
  -lr LEARNING_RATE, --learning_rate LEARNING_RATE
                        initial learning rate
  --undirected UNDIRECTED
                        whether it is an undirected graph
  -a ALPHA, --alpha ALPHA
                        the rate of structure loss and orth loss
  -l LAMB, --lamb LAMB  the rate of structure loss and guilded loss
  -g GRAD_CLIP, --grad_clip GRAD_CLIP
                        clip gradients
  -K K                  K-neighborhood
  --sampling_size SAMPLING_SIZE
                        sample number
  --seed SEED           random seed
  --index_from_0 INDEX_FROM_0
                        whether the node index is from zero


If you find this code useful, please cite our paper:

  title={Deep recursive network embedding with regular equivalence},
  author={Tu, Ke and Cui, Peng and Wang, Xiao and Yu, Philip S and Zhu, Wenwu},
  booktitle={Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery \& Data Mining},