I have a large dataset to run a specific Graph Data Science algorithm on.
The functional requirement is that the algorithm will be run often and that the dataset changes in real-time.
As I understand, in order to run an algorithm I have to project the persistent graph into memory first.
But, GDS only provides a projection of the whole dataset once (as a (filtered) snapshot), therefore, on each change to my dataset (i.e. a new relationship edge added between two nodes), I have to rerun the projection again, which seems quite an ineffective thing to do.
Is there a generic way to circumvent this and keep the Projection properly in sync with the persistent graph?
