Graph databases do not support subgraphs or multilayers, which will require a programmatic abstraction to implement over the database.
- Classifying graphs into groups that may overlap.
- Layering graphs allowing connections between them, kind of aggregating connections according to specific needs.
- multi-layer views (Subgraphs) of a target unit/node.
- Changing representation of the graph according to the dimension model used with the target node. e.g. timescale dimension, where the target node is and entity and restricted to events relationship.
- representing multiple units in relations to each other, or classified by one another. Diffenent algorithms to merge nodes into view: mutually inclusive, nested / included by one another, side by side, etc.
- Cross relationships can be applied between the graph layers.
- Allowing to restrict to a target relationship types.