This repository contains the implementation of paper "NovTrack: Dynamic Cluster Analysis for Tracking Novelty in Network Telescopes"

1. Clone this repository
2. Create and activate a virtual environment (recommended):

cd novtrack
python -m venv ./.venv
source ./.venv/bin/activate

3. Install Python dependencies:

pip install -r requirements.txt

• data/parsepcap.ipynb provides function to parse .pcap file with dpkt
• You can follow the steps to train sender embeddings (generate_embeddings.ipynb), cluster the senders (run_clustering.ipynb) and track cluster changes ((run_tracking.ipynb))

Our dataset is available upon request, please contact the authors to get access