https://github.com/wallarm/wallnet
Open-source code to support BSides 2019's talk: Bye-Bye False Positives: Using AI to Improve Detection
https://github.com/wallarm/wallnet
cybersecurity machine-learning tensorflow
Last synced: about 1 year ago
JSON representation
Open-source code to support BSides 2019's talk: Bye-Bye False Positives: Using AI to Improve Detection
- Host: GitHub
- URL: https://github.com/wallarm/wallnet
- Owner: wallarm
- License: mit
- Created: 2018-09-10T15:52:03.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2024-01-19T17:19:01.000Z (over 2 years ago)
- Last Synced: 2024-03-26T19:07:45.155Z (about 2 years ago)
- Topics: cybersecurity, machine-learning, tensorflow
- Language: Python
- Homepage:
- Size: 19.5 KB
- Stars: 19
- Watchers: 7
- Forks: 5
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# WallNet
(Released a part of the BsideS Talk 2019: https://www.youtube.com/watch?v=C4yJR35tAhg. OSS project is not maintained)
WallNet is an bidirectional recurrent neuralnetwork with attention mechanism and pooling layers and pipeline for Structured Query Language injections (SQLi) detection. To illustrate the application of this methodology, we will review in detail the implementa-tion of AI-based false-positive detection for a SQL injection. WallNet developed on [TensorFlow 1.11](https://github.com/tensorflow/tensorflow/releases/tag/v1.11.0) and Python3.6.
This implementation is an baseline for [Malicious Intent Detection Challenge](https://www.kaggle.com/c/wallarm-ml-hackathon)
## Build
Firstly install dependences:
```
apt install -y swig
apt install -y python3, python3-dev, python3-pip
pip3 install -r requirements.txt
```
Now you have to build project.
```
./build.sh
```
---
## Using
### Preparing dataset
```
python3 data_loader.py --dataset_info_file=./data/train.csv --dataset_file=./data/train.msgp
python3 data_loader.py --dataset_info_file=./data/test.csv --dataset_file=./data/test.msgp --test
```
For more information use help: ```python3 data_loader.py --help```
### Train
```
python3 train.py
```