Using a graph approach makes sense when your data and your questions involve connections. In some datasets, the connections are as important as the individual entities. In a money laundering investigation, for example, it is crucial to capture how money flows between individuals and companies. Similarly, some questions are particularly well suited for graphs: “how X and Y are connected”, “what is X connected to”, “what’s the role of X in the network”. The world biggest companies have been relying on graphs for years now with systems such as Google’s “Knowledge Graph” or Linkedin’s “Enterprise Graph”.
Among the use cases in which the graph approach is the most popular are cybersecurity, anti-financial crime or intelligence analysis. In these domains, the organizations switching from tables to graph benefit from:
- a unified view of their data instead of blind spots and silos;
- the ability to run complex queries without hitting performance bottlenecks.
A more complete data picture and the ability to detect complex patterns are invaluable assets to identify cases of fraud or other threats in large volumes of data. For the banks, government agencies and other organizations turning to a graph approach, it leads to the discovery of new threats and faster investigations.