ML for IoT vulnerability detection program

This project is a Machine Learning (ML) based solution to detect potential security risks in Internet of Things (IoT) software.

Run the machine learning pipeline:

Once required packages were installed, run the command to extract the vulnerability database from the given input projects as listed in config/classifier.yaml:

python3 -m classifier.classifier

Command to run the classifier with overriding model, multiclass and granularity-level parameters.

python3 -m classifier.classifier --model LSTM --type multiclass --granular function

Additional script:

Script can be executed before classifier.py, to make <data>.csv file available for the training, testing and the evaluation of the machine learning and deep learning models. The script get statement and function tables of the IoTvulCode database and save as data/IoTvulCode/IoTvulCode-statement.csv and data/IoTvulCode/IoTvulCode-function.csv.

python3 -m classifier.db2csvs -d data/IoTvulCode.db -o data/IoTvulCode/

Configurations:

The config/classifier.yaml lists the configurations for the extraction the vulnerability data from the given projects: list, for example:

Note: Run all the commands from the parent directory of the repository IoTvulCode.