In this project, we open-source the source code and data sets of our Learnable Monotonic Vertex Embedding (LIVE) approach for exact subgraph matching.
- The codes require the following dependences:
- A modern C++ compiler compliant with the C++17 standard (gcc/g++ >= 12.2)
- CMake (>= 3.28)
- Under the Offline directory, execute the following conda commands to configure the Python environment.
conda create --name <new_environment_name> --file requirements.txt
conda activate <new_environment_name>
- Turn into the Offline directory, execute the following command to train the embedding model.
| Parameter |
Default Value |
Description |
| -n |
../Dataset |
Dataset Path |
| -d |
2 |
Embedding Dimension |
| -e |
1000 |
Epochs |
| -b |
4096 |
Batch Size |
| -l |
0.01 |
Learning Rate |
- Turn into the Online diectory, execute the following command to build the project.
mkdir build
cd build
cmake ..
make
- Return to the root directory, and execute the following command to run a quick start example.
| Parameter |
Default Value |
Description |
| -d |
../Dataset/ |
Dataset Path |
| -q |
../Dataset/query_graph.graph |
Query Graph Path |
| -a |
1000 |
Alpha Value |
| -b |
0.01 |
Beta Value |
| -e |
2 |
Embedding Dimension |
| -k |
2 |
Hop Number |
