Understanding the importance of graph data modeling
Before we go ahead with looking at how GraphRAG flow works with Neo4j, let us take a step back and understand how we can model knowledge graphs. We will take some simple data and try to look at how we model that data in RDBMSs and graphs. We will also see how this modeling differs depending on how we see that data.
Graphs force us to think in different ways and see the data from different perspectives depending on what we are trying to solve. While this might seem like a problem, it is actually opens a lot of doors. For a long time, we have been taught to think of the RDBMS storage approach in terms of Entity-Relationship (ER) diagrams. This approach was good for representing/persisting data when there were limitations in the technology, and storage costs were very high. With technologies evolving and hardware becoming cheaper, new avenues have opened and new approaches to model data are possible. Graphs are well suited to take...
