If the project you are working on generally contains many relationships and parent-child relationships, it makes sense to use it to improve your search performance. However, we don't just use Neo4j, we build a hybrid structure with RDMS in terms of performance where needed.

It is very useful in query speed in large-scale and multi-correlated data.

It requires learning new notations compared to the usual databases and the consultancy service is limited.

Overall, Neo4J is a great product with which it is possible to implement many aspects of the field of data science. Especially the great support of the developers is worth mentioning. Neo4j also offers the possibility to perform optimized operations on the data and to visualize the results directly. This eliminates the tedious part in the data science area, which deals with planning and setting up an infrastructure.

Neo4J can be used for various use cases. Thereby, Neo4J can not only be used in the backend for use cases. Rather, Neo4j offers the possibility that scientific projects can also be realized on already pre-programmed algorithms and via the Neo4J browser or via Neo4J Bloom. Thereby it is convincing that Neo4J can be almost an own IDE for Data-Science and still has the possibilities of a database. This creates new possibilities for scientists as well as for business applications to run optimized data analysis directly on the database.

It is contradictory that Neo4j has training programs on the one hand and is open to the data science community, but does not make the really important aspects, such as data security through server-side certification, easy to understand or role-based access rights accessible. Such features are the essentials when it comes to secure data management.

Its best suited for the performance it provides for the complex databases

UI performance for visual node graphs can be better

I deployed Neo4j as a NoSQL database, more specifically as a graph DB. I am using this product to create my graph which will be generally composed of nodes as well as links to relate them. Neo4j provides a very impressive graph visualization, as well as a powerful querying tool. It should be noted that even the query results are graphically illustrated. Several libraries/drivers in many languages (i.e., python, java ...) exist for interacting with Neo4j.

This products works very well as a graph database. It satisfactorily represents the nodes and their links allowing the users to navigate and interrogate them. However, when we try to visualize a very complex graph, it is not as efficient.