Developing the world's first transnational open beneficial ownership register: A brief history of the Open Ownership Register
Data partnerships and experiments
As well as developing the OO Register to make best use of publicly available BODS data, Open Ownership has leveraged the data powering it to pursue a range of partnerships and experiments to showcase how structured and interoperable BODS data can be used in multiple ways, which are outlined below.
Using BODS data in graph database or linked data formats
OpenScreening (February 2023)
OpenScreening: A Neo4J-Powered Free Compliance Investigation Tool
To demonstrate how to uncover hidden relationships between beneficial owners, politicians, sanction targets, and networks of companies, we collaborated with OpenSanctions and Linkurious to launch OpenScreening. This proof-of-concept tool takes BO data from the OO Register and creates a graph database to match beneficial owners to names appearing in sanctions data from OpenSanctions, or in leaked BO data from the ICIJ. This database was then uploaded and made available to explore via Linkurious thanks to its Neo4J graph database format. Check out a webinar from Linkurious to learn more.
Black Ice’s SARA (July 2023)
We collaborated with Black Ice AI to showcase how to analyse BODS data using their Suspicious Activity Risk Awareness (SARA) tool. This used Senzing’s entity resolution technology to bring together data from multiple sources (including the OO Register) to resolve entities and relationships. SARA’s platform can incorporate databases such as the ICIJ’s Offshore Leaks, the Dow Jones Watchlist, Moody’s Orbis, OpenCorporates, and AuthID. This collaboration also generated important insights for the Open Ownership Technology Team into the number of duplicate and related records in the BODS data.
The value of resolving Open Ownership data inside SARA
BODS risk detection (September 2023)
Learning which companies have been awarded public contracts
Combining high-quality BO data with other datasets can be crucial to help detect potential risks during customer due diligence, know-your-customer checks, and sanctions screening processes. To demonstrate the value of our BODS RDF vocabulary, we showed how once in this format BODS data can be combined with public procurement data published in line with the Open Contracting Data Standard as well as with sanctions data published by OpenSanctions using the FollowTheMoney data model and the Offshore Leaks database from the ICIJ. The resulting data was then queried using RDF/SPARQL (a query language used to express queries across diverse data sources) to leverage its graph nature for a series of risk and compliance use cases. Both individuals and companies can be treated as targets.
GraphAware webinar (May 2024)
Unlocking Complex Ultimate Beneficial Ownership Investigation
During a joint webinar, the GraphAware team loaded BODS data from the OO Register into Hume, its graph data visualisation and exploration tool. This information was then combined with sanctions data and enhanced using entity resolution to demonstrate how it can support detailed investigations by law enforcement or analysts.
Data analysis and querying BODS data
As well as partnering with external organisations, Open Ownership worked with Open Data Services to demonstrate the range of data analysis approaches which can be used to generate deeper insights from BODS data.
Leveraging the BODS data analysis tools, we created a reproducible notebook with UK PSC Register data to demonstrate how to write queries to spot a number of red flags using BODS data.
The questions we sought to answer were:
- How many entities declare that they have no beneficial owners or that their beneficial owners are unknown?
- How many entities have ownership networks which involve natural persons or entities in countries on the European Union list of non-cooperative tax jurisdictions?
- How many entities have complex ownership chains? (Defined here as five or more observable layers of ownership.)
- How many entities were founded and subsequently dissolved within a year?
For example, we learned that there were 71,774 companies with ownership chains involving five or more levels that could be observed in the UK data at that time.
Increasingly, tools are being created to make it easier to use, query, and process data files in formats such as Parquet without the need to download data or install a database. One such tool is DuckDB, a fast and efficient online analytical processing structured query language (SQL) database management system, which only requires minimal setup.
To demonstrate the capabilities of DuckDB for querying BO datasets hosted by Open Ownership – once we had added the Parquet export format to our BODS data analysis tools – we created a reproducible notebook showing a number of queries which create data samples or list people who have direct BO interests in a company.