Graph databases and network data model

Question

Database

How do they relate to each other? Are graph databases an implementation of the network model?

For some reason, the English wiki states that graph databases are a general case in contrast to a specialized network model: "General graph databases [...] are distinct from specialized graph databases such as [...] network databases." ( graph database )

Answer 1

The network, depending on the task, is an undirected/directed/bidirectional graph. For example, an artificial neural network (ANN) of direct propagation (Feedforward) is a directed graph (neural impulses go in one direction), and a recurrent ANN is bidirectional (an error signal goes back).
Usually in graph theory, a network is a weighted directed graph, i.e. a graph in which each connection is associated with a certain number. These numbers estimate the "cost" of the path along this link, or the length of the link, as on a road map. In each specific case of using a graph as a formal means of describing a problem, these numbers can be interpreted in their own way.
“In the programming world, there is no consensus on which of the two terms is “graph” or “network”. We chose the term "network" because it seems to be more common in application areas.
See Graph Theory Glossary
Graph databases are optimized to store such structures. As for the "implementation of the model", you can say so. It is also possible to store in MyISAM/InnoDB, but specialized engines, such as OQGraph , are better suited for this .

Answer 2

As far as I know, in a graph data model, there is always only one main ancestor, from which the rest of the nodes come. The scheme will be similar to a tree, from the root of which individual branches emanate.
In the network model, an arbitrary node can have multiple independent ancestors. Those. if you try to draw an abstract diagram, you get a semblance of a sea network.