The use case for the graph database is to have users and contents (vertices) linked by likes, favorites and reports relations (edges). The problem I have is that I will sometimes need to show the reported contents (from any users). Since this is not a standard graph traversal, I fear this would have a big performance hit.
Is it possible to index the edges of type "reports" to quickly get the list of all contents that have been reported? Is there a better way to do this?
No, you cannot (don't need to) explicitly manage indices. Neptune uses a novel indexing strategy based on semi-clustered indices and offers excellent index performance out of the box. There is no need for custom indices.
From Neptune FAQs: https://aws.amazon.com/neptune/faqs/
Do I need to create indices on my data with Amazon Neptune?
No, existing graph database users are often forced to try and outguess the vendor implementation. Explicitly maintaining indices is just one aspect of that. Amazon Neptune does not require you to create specific indices to achieve good query performance, and it minimizes the need for such second guessing of the database design.
Can you share some details on the specific queries that you are looking for?
