Hey all!
A couple of months ago, I wrote up a little sketch for a new, configurable loading tool that would put followthemoney data into Neo4J. The idea was to make the process more configurable, and load straight to the database using bolt and query batching.
The vibes and I have had a weekend of it, so there’s now a little draft repo:
It’s not on PyPI yet, but works reasonably fast and has a lot more flexibility. The demo/ directory shows how to apply it to a set of random FtM datasets.
There’s parallel development underway from C4ADS here: followthemoney-neomodel/followthemoney_neomodel.py at main · C-Research/followthemoney-neomodel · GitHub - basically using an ORM to map FtM entities into Neo. Very cool, and unlike ftmg it has the potential to programmatically create Cypher queries in Python!
Cheers,
- Friedrich