Ep. 183 Hidden Relationships in Large Data Sets

John Gilroy on LinkedIn   https://www.linkedin.com/in/john-gilroy/

Want to listen to other episodes? www.Federaltechpodcast.com

History books will document the origin of the relational database at around 1970. About a decade later graph technology was introduced but it has taken decades for the cost of storage to go down and the ability to compute to go up. Finally, we can take advantage of a new way to unlock answers from a database.

A typical relational database looks at information in tables. This can be fantastic for many actions, which is why it became popular. However, drilling down into information can involve re-indexing and hopping around tables.

Graphing technology looks at the data and tries to find relationships. As consumers, we know if we purchase an expensive couch with a credit card, the credit card company may email and question if that is a valid purchase.

Well, multiply that by hundreds of thousands of users and millions of data points. It is not just a couch; it may be automated financial transactions that involve fraud.

Attend the Neo4j Graph Summit Government event on October 9th at the Spy Museum in Washington, D.C.

For a human to sit down with some tables of data would make the process so time-consuming, that millions could be stolen before the culprit was discovered.

During the interview, John Bender from Neo4J explains how they respect the existing data structures but can layer on a deeper understanding of the relationship between a specific transaction and an outcome. In other words, you will not have to say goodbye to your data silos.

Another application is understanding the supply chain. Because so much hardware and software are outsourced, it is hard to connect the dots. John Bender refers to an Army project where they have eight million nodes and twenty-one million relationships.

Listen to put into perspective new ways to improve analytical speed and reduce risk from fraud to the supply chain.