DataSynthesizer

What DataSynthesizer is

DataSynthesizer is an open-source tool for generating synthetic data that simulates a supplied dataset. The project says it applies differential privacy techniques to provide a privacy guarantee. Its demos cover random, independent-attribute, and correlated-attribute modes. The documented input must be a table in first normal form, so this requirement shapes which datasets fit the tool. You can install it with `pip install DataSynthesizer`, run example notebooks with Jupyter Notebook, or launch the included Django web interface locally and open it in a browser at `/synthesizer`. The project aims to facilitate collaboration between data scientists and owners of sensitive data. Its identified integrations are Python pip, Jupyter Notebook, and the companion Django interface. The repository identifies the license as MIT and provides GitHub Issues for questions and reports. DataSynthesizer is free and self-hosted, with web use through the locally run interface. The listed scope excludes relational and unstructured data.

Who it suits

DataSynthesizer may suit data scientists working with owners of sensitive data when the input is a first-normal-form table. Users can work through notebooks or run the Django interface locally.

Limits

The free software requires input in first normal form. Its stated scope excludes relational and unstructured data.

Getting started

1. Install the package with `pip install DataSynthesizer`.
2. Prepare the input as a table in first normal form.
3. Run an example notebook with Jupyter Notebook, or start the Django project locally and open `/synthesizer` in a browser.
4. Use GitHub Issues for project questions or reports.

Verdict

DataSynthesizer provides several modes for generating synthetic data from tabular input, with notebook and local web workflows. Its documented scope excludes relational and unstructured data.