Fraud Graph Network

Graph databases can help us better understand and detect fraud patterns. Lets play around with Graph databases and algorithms and see if we can come up with a creative way to build an interactive fraud network

In this project we will use the neo4j graph database. What needs to be done on a high level is the following:

develop an initial graph model, i.e what are the nodes (entities), and the relationships between them --> this includes use cases and questions we want to answer, e.g which fraudulent emails are connected to the same ip address ?
set up a neo4j database
import the data from snowflake to the neo4j database
explore different use cases/ questions by queuing the graph using Cypher
explore if we can use graph algorithms such as Louvain to detect fraudulent patterns