This repository contains implementation of DHNE : Network Representation Learning Method for Dynamic Heterogeneous Network.
DHNE combines the historical information into current information in the netwrok to learn the representations of nodes in dynamic heterogeneous networks .
- python 3.4 (or later)
- networkx 1.11
- gensim 2.3.0
Please use --dataset argument, where dataset-name can be one of the following: "Dblp", "Aminer".
cd code
python DHNE.py --dataset Aminer
The output will be saved in /Aminer/aminer_result folder
We experiment on two real-world datasets: DBLP, Aminer datasets
- Folder "Dblp/dblp_dataset" contains DBLP dataset graphs. There are 19 graphs from 2000 to 2018 .
- Folder "Aminer/aminer_dataset" contains Aminer dataset graphs. There are 16 graphs from 1990 to 2005.