How does a graph database differ from a relational database?

The kind of information that we’re going to look at that we’d like to be able to represent and look at is some entity A since some money to entity B and send some money to entity C an entity B sends it to these three other guys part of the money goes to each of them an entity C sends some money to these two other guys and it all comes back to the originator it all comes back to entity A so if we saw a pattern like that we might suspect that money-laundering is going on this guy sends out money and it all comes back to him now if you had a relational database with these little tables you’d have to start looking in this table that is a this account and then you’d have to say that he’d wrote a check to be a need to have to go over and B’s account and look there and and you’d be searching and table after table and relational databases can eventually do that but they’re terribly inefficient at it to the point that it would take so long people don’t even bother to try.

In a graph oriented database we actually represent the data as something equivalent to these little circles and arrows you think of the circles if it representing some noun this is a company or it’s an individual with a bank account or it’s a certain person and in the and here’s some other down here on this circle and there are some verb represented by the arrow this person lives in Seattle or this person owns a Toyota or something like that so it’s always noun verb noun and all the data in the database is stored that way and it has this nice property that if you storing these triples of information they correspond exactly to what we think of mathematically as a directed graph and which is just a set of these circles then they’re connected by by arrows so conceptually this is the kind of data that stored in a graph oriented database as opposed to a relational database and it does well at searching for certain patterns in this graph structure

Similar Posts: