How graph databases work

Relationships

Conclusions

Where the task in hand is primarily a matter of processing statements about relations between people, things, events, points in time, or places, graph databases are superior to their relational siblings for several reasons. Queries for relationship chains in particular, such as my friends' friends, need recursive joins in the relational world, which are very complex and slow and become all the more evident the deeper the concatenation goes. Graph databases are orders of magnitude more powerful for such problems. NoSQL databases like key-value stores or document-oriented databases have similar disadvantages. These databases often shift the data processing onus to the application when relationships are queried.

Another advantage of graph over relational databases is that they can be extended by new relations at any time and do not require a pre-defined schema. This circumstance also makes them more agile because they adapt better to a step-by-step expansion of the program logic. Furthermore, relational schemas, such as those that can be designed in an entity relationship model, only support simple and unidirectional relationships, whereas in a graph database, they have specific properties and can be bidirectional.

The opportunities to capitalize on these advantages are constantly growing as the world becomes increasingly interconnected. The value of data lies not only in sheer volume, but in the connections, and graph databases are the means of choice for representing, storing, and querying these connections .

Infos