dp-sequential-events 1.6

Privatized

Differential Privacy in Sequential Event Logging

---

✨ Project Description

Sequential event logs often contain sensitive information. dp-sequential-events implements differential privacy (DP) techniques to anonymize sequential event logs while preserving statistical properties for analysis.

The pipeline follows these steps:

1. DAFSA annotation of event logs
2. Filtering based on probabilistic risk measures
3. Differentially private case sampling
4. Laplace noise injection for timestamps
5. Reconstruction of anonymized timestamps
6. Final privacy-preserving event log generation

---

🗂 Repository Structure

dp-sequential-events
┣ 📂 src
┃ ┃
┃ ┗ 📂 dp_sequential_events
┃   ┣ 📂 main
┃   ┃ ┣ main.py
┃   ┃ ┣ annotated.py
┃   ┃ ┣ filtered.py
┃   ┃ ┗ case_sampling.py
┃   ┗ 📂 databases
┣ pyproject.toml
┗ requirements.txt

🚀 Online Execution

You can run the CLI in Google Colab or locally. For Colab: Open in Google Colab

Install from PyPI:

pip install dp-sequential-events

Run the CLI tool:

privseq

👩‍💻 Authors

Marta Jones 💻

Anailys Hernandez 💡