0

I am working on a metadata search project in which I have a graph database. The graph database is created by combining multiple datasets which contain information about users (purchase history, CRM data, etc.) and contains information like user name, age, date of birth, email, SSN etc.

So for a single user, there could be multiple instances in the database(one from CRM, one from purchase history, etc.). I want to build a search system that should take some user metadata as a query and return information combined from all instances of a user.

Is resolving records online(while executing a query instead of previously resolving all the results) is a Good Idea? Is a Graph Database like neo4j a good choice for this task? If yes, what is the optimal way of achieving this in neo4j?

Yash Thenuan
  • 591
  • 5
  • 17
  • You would probably want to have a single node representing each customer and all the metadata connected to it instead of multiple instance of the same customer – Tomaž Bratanič Oct 20 '21 at 06:31
  • Yes exactly, that is one way of doing things, was just curious if figuring out that part(single customer node instead of multiple) is possible on runtime. – Yash Thenuan Oct 20 '21 at 12:54
  • It's possible, but not optimal... depends on your customer identifiers – Tomaž Bratanič Oct 20 '21 at 13:03

0 Answers0