POLE Investigations with Neo4j
Download as PPTX, PDF•4 likes•2,715 views
This document discusses using Neo4j and graph algorithms for police investigations (POLE) based on a real-world crime dataset from Greater Manchester. It introduces the POLE data model and use cases, demonstrates Neo4j's advantages over other NoSQL databases for relationship queries, and shows sample visualizations of the crime graph data in Tableau and a custom web app. The presentation concludes by discussing ways to extend the demo dataset and tools to support a full-fledged investigative analysis platform.
POLE Investigations with Neo4j
- 1. POLE Investigations using Neo4j and Graph Algorithms Joe Depeau Sr. Presales Consultant, UK 20th March, 2018 @joedepeau http://linkedin.com/in/joedepeau
- 2. • What is the POLE Data Model? • Why Neo4j? • Neo4j POLE Demo • Sample POLE Data Visualisations • Extending the Demo for Real-world Use • Q & A 2 Agenda
- 3. What is the POLE Data Model? 3
- 4. 4 The POLE Data Model Vehicles Evidence Weapons Documents Emails Phones Victims Suspects Witnesses Investigators Employers Family Members Crimes Arrests Meetings Data Transmissions Phone Calls Interventions Crime Scenes Home Addresses Places of Employment Public Buildings Landmarks Travel Destinations Objects Persons Events Locations
- 5. • Policing • Counter Terrorism • Border Control / Immigration • Child Protection / Social Services • Missing Persons • Prisoner Rehabilitation 5 POLE Use Cases Real Time Proactive Reactive Insights
- 6. 6 But … what about Big Brother?!
- 7. Why Neo4j? 7
- 8. 8
- 10. Blank SlideUsing Other NoSQL to Join Data Using Neo4j Slow queries due to index lookups & network hops Lightning-fast queries due to replicated in- memory architecture and index-free adjacency Relationship Queries on non-native Graph Architectures MACHINE 1 MACHINE 2 MACHINE 3 UNIFIED, IN MEMORY MAP 10
- 12. • UK street-level crime data is freely available from data.police.uk • We will be looking at street-level crime data from the Greater Manchester Police for the month of August 2017 • The crime data provides unique crime IDs, longitude and latitude (at street or ‘block’ level), month, crime type, and last outcome • The crime data does not include personal identifiers (not even anonymised tokens) • Longitude/latitude values were translated to UK postcodes using public APIs • Random data was generated for people, officers, phone calls, crime date, etc. • The crime and random data where combined and curated to create the demo 12 About the Demo Dataset ○ Locations: 14,904 ○ Crimes: 28,762 Relationships: 105,853 ○ Officers: 1,000 ○ Persons: 368
- 15. • We’ll view a few example visualisations created using Tableau Public: • A geographic representation of crimes in the database • A chart of crimes by type and date • A geographic representation of the centrality algorithm results • Connectivity between Neo4j and Tableau Public is managed by the Neo4j Tableau Web Data Connector v2.0 • Demonstrates the types of Geospatial and BI visualisations that can be designed on top of a POLE graph 15 Sample visualisations using Tableau
- 16. • We’ll also view an example front-end using Neo4viz, an internally developed tool for creating visualisations. • Demonstrates how an end-user POLE application interface might look. • Neo4viz was developed using: • SpringBoot • ZK Server • Font Awesome & Ionicons • vis.js 16 Sample visualisations using Neo4viz Neo4j POLE data SpringBoot Web App Browser App
- 17. Extending the Demo for Real-world Use 17
- 18. • Using ‘Personas’ instead of ‘Person’, to account for things like aliases. • A richer set of relationships between Persons and Crimes (i.e. Witness_To, Victim_Of, Suspected_Of, Convicted_Of), Locations (i.e. Works_At, Visited, etc.), and Objects (i.e. Owner_Of, Driver_Of). • Supporting traceability and auditing of data. In real life it’s very important to understand the lineage of the data (who entered the information and when, who updated it, has it been verified, etc.) and how we could demonstrate we have the right to hold that information (i.e. was it discovered as part of an investigation, is it publicly available, etc.). • A robust security configuration, to restrict data access to those who have the right authorisation. • Adding weighting to our searches and algorithms - for example some crimes might be considered more dangerous than others (i.e. Violence and Sexual Offences are more serious than Shoplifting), or some relationships might be considered closer (i.e. ‘Family’ or ‘Lives With’ may be weighted more than ‘Social Network’).18 Ways the demo could be extended
- 19. Q & A 19
- 20. Thanks for your time! (the end) 20
- 21. • Open data about crime and policing in England, Wales, and Northern Ireland: http://data.police.uk • Neo4j Tableau Web Connector: https://github.com/neo4j- contrib/neo4j-tableau • Neo4j Graph Algorithms: https://neo4j.com/developer/graph- algorithms/ 21 Links
Editor's Notes
- #19: _