followthemoney 4.5.3

A data model for anti corruption data modeling and analysis.

Follow the Money

This repository contains a pragmatic data model for the entities most commonly used in investigative reporting and financial crime investigations: people, companies, assets, payments, ownership relations, court cases, etc.

The purpose of this is not to model reality in an ideal data model, but rather to have a working data structure for researchers. Complex legal considerations are simplified to allow for efficient data processing.

followthemoney also contains code used to validate and normalize many of the elements of data, and to map tabular data into the model.

Documentation

For a general introduction to followthemoney, check the high-level introduction:

• https://followthemoney.tech

Part of this package is a command-line tool that can be used to process and transform data in various ways. You can find a tutorial here:

• https://followthemoney.tech/docs/cli/

Besides the introductions, there is also a full reference documentation for the library and the contained ontology:

• https://followthemoney.tech/explorer/schemata/

There's also a number of viewers for the RDF schema definitions generated from FollowTheMoney, eg:

• LODE documentation
• WebVOWL
• RDF/OWL specification in XML.

Installation

You can install followthemomey via PyPI like this:

pip install followthemoney

The most tricky dependency of followthemoney is pyicu, which helps us to perform certain text operations. Please review their documentation if you run into any issues with ICU.

Development environment

For local development with a virtualenv:

python3 -mvenv .env
source .env/bin/activate
pip install -e ".[dev]"

Now you can run the tests with

make test

Releasing

We release a lot of version of followthemoney because even small changes to the code base require a pypi release to begin being used in downstream applications. To this end, here's the steps for making a release:

git pull --rebase
make build
make test
git add . && git commit -m "Updating translation files"
bumpversion patch
git push --atomic origin main $(git describe --tags --abbrev=0)

This will create a new patch release and upload a distribution of it. If the changes are more significant, you can run bumpversion with the minor or major arguments.

When the schema is updated, please update the docs, ideally including the diagrams. For the RDF namespace and JavaScript version of the model, run make generate.