Implementation of TBCNN using Tensorflow. Our report here.
Instead of training on C programs, This version is used for SQL injection detection. Dataset is provided in the data folder.
It's highly recommended to install into a virtual environment. Run the following command after activating venv:
pip install -r requirements/dev.txt
tbcnn.embedding is the entry point for pretraining the embedding matrix.
tbcnn.tbcnn is the entry point for the tbcnn model. It requires a pre-trained embedding matrix.