LouvainNE: Hierarchical Louvain Method for High Quality and Scalable Graph Embedding
The 'model' folder contains the codebase for generating the HCNE embedding vectors for nodes in an input graph. To generate the embedding vectors, run the following command inside the 'model' folder to generate embedding vectors for nodes in the karate club network:
./louvain karate.txt -a 0.005 -n 20
where karate.txt -> edgelist file for the karate network -a 0.005 is the alpha parameter for combining the embeddings -n 20 sets the dimension of the embedding vectors to 20
The 'Evaluation' folder contains the code for evaluating generated embedding vectors on different downstream prediction tasks as well as generating a sampled graph representative of the original graph.
For our experiments, we are using the following datasets:
Blogcatalog: http://socialcomputing.asu.edu/datasets/BlogCatalog3 youtube: http://socialnetworks.mpi-sws.org/data-imc2007.html flickr: http://socialnetworks.mpi-sws.org/data-imc2007.html