Hi , So I want to use neo4j as my graphical DB(to store my nodes and edges) and also as vector DB . I need the vector DB to find the simialrity between my nodes(i.e for a given node find the other related nodes) and then these related nodes should be connected via edges nd this edge info should be stored in neo4j graph db.
SO can we do this?
can we first store our nodes (no edge details ) and store their embeddings(do we store it in same neo4j or maintain separate one) and then query the embeddings to find the relation between the nodes ??
Typically, we store the embeddings in the graph as a property on an entity (typically a node), and then you can design your data model in the graph to add relationships where needed.
Start with everything in the same graph. Perhaps if you ended up with a really massive graph, you might need separation, but that’s likely the exception rather than the rule.