Organizations, especially a new class of employees like business analysts, data scientists, and marketing strategists, are actually trying to get better insight using the relations in all the data records of their enterprise. In order to better serve these new ways of searching for information, it is becoming critical to find a way to expose existing enterprise data in a graph database format. Unfortunately, most of today’s data is stored in relational database management systems (RDBMS).
In the relational data storage model, there is no physical link to indicate the relation among different data records. They are only linked logically by a special attribute called a foreign key, rendering the search of related records not as effective. In order to find relationships in a relational data storage model, you have to build complex queries that perform actions over and over again. It’s not very efficient. While relational databases are widely used, they are not ideal for storing and querying data that have a high degree of relationships.
On the other hand, the graph data storage model keeps data attributes together naturally (in nodes). By creating physical links called edges, all related nodes are kept together, highlighting the relationships between data sets. This focus on relationships in graph database makes it a far better home for today’s interconnected data than traditional relational databases.
But, since most of today’s data is still in relational database format, how do you convert your current data to a graph database format? Mapping data from a relational database to graph database is fundamentally a task of converting the relational representation from one database to the other. More specifically, we can use the foreign keys of relational data model to build edges, thus transforming loosely coupled data records into a highly bounded group of nodes. Nodes form the basis of graph database architecture.
For a detailed walk-through of how to do this, download the brief: “How to Map Relational Data to a Graph Database”. In it, we go step by step and use a well-known example to demonstrate an approach to map information from a relational database to a graph database.