Searching for Shady Patterns: Shining a light on UK corporate ownership
summary
Built a graph database mapping the UK beneficial-ownership register (about 4.5 million companies and 4 million people), ingesting Companies House dumps plus officer data, cleaning messy free-text fields and deduplicating identities. Uses Python, pandas, AWS and a Neo4j backend with a Neo4j sandbox for demos. Finds thousands of under‑age owners, postcode clusters with 20,000 companies, cyclical control chains, and produces lead lists used by journalists and Companies House investigations.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 3:44UK corporate transparency
- 8:36Corporate ownership graph
- 12:29Neo4j sandbox demo
- 16:04Data modeling challenges
- 20:00Project impact
- 26:52Detecting suspicious patterns
- 31:49Key findings
- 33:37Conclusion
related talks
42mFighting fraud with graph databases
24mGuardian - fraud detection using graph analysis
32mCode More, Draw Less, and Debug Just a Little!
14mDatashare + Neo4j: Unstructured Docs to Knowledge Graphs for Investigative Journalism
30mInterconnected Capital: Visualizing WAM, PE, RE, and HF Data with Neo4j
20mPower Amazon Bedrock applications with Neo4j knowledge graph
40mThe Paradise Papers: Behind the Scenes
29mFrom Warehouse to Wisdom: Activating Business Data with Neo4j