One of the big use cases of Neo4j is fraud detection. From discovering identity fraud and skimming locations to insurance fraud and visualizing networks, putting your stuff in the graph allows you to focus on the suspect cases.
In most cases, investigation is carried out by teams of specialists. Each team member is focusing on a number of cases and almost automatically memorizes names and numbers that pop-up, assuming that bells start to ring whenever a new fraud happens. There is of course quite a big risk of not seeing how individual fraud cases are connected or missing them all together, not just because different team members have memorized different names, but also because of the increasing complexity of cases and criminal setups.
By using Neo4j a graph database in combination with our InterActor dashboard, the task becomes easier. Networks can be visualised with a click on a button, items like accountholders and IP addresses can be added, and when identities seem fraudulent, aliases can be created in a way that they pop-up immediately when another team-member finds items that are common between cases.
This demo was built with InterActor Enterprise, using Neo4jQuery, Networkview and TableView (:Function) nodes, and connected by [:TRIGGER] relationships that define the interactions. The picture below contains the entire demo application, built in less than 3 hours.
Docker
Cloud