liveBook · Manning

Homogeneous graph

As explained in Graph Neural Networks in Action

A homogeneous graph is a graph containing one type of node and one type of edge.

A homogeneous graph contains one type of node and one type of edge. Use this model when every entity and relationship shares the same role, such as candidates connected by relationships in a recruitment network. section 1.2.2

Distinguish it from related graph structures:

  • A heterogeneous graph expands the model with multiple node or edge types, such as candidates, companies, and employment relationships. section 1.2.2
  • A bipartite graph partitions nodes into subsets and permits connections across subsets, not within the same subset. section 1.2.2

Recruitment network

Represent job candidates as nodes and relationships between candidates as edges. Adding companies and employment links changes the model into a heterogeneous graph. section 1.2.2

Figure 1.2 A homogeneous graph and a heterogeneous graph. Here, the shade of a node or edge represents its type or class. For the homogeneous graph, all nodes are of the same type, and all edges are of the same type. For the heterogeneous graph, nodes and edges have multiple types.
figure

Compare uniform node and edge types with the varied shades representing types in a heterogeneous graph. section 1.2.2

How Graph Neural Networks in Action covers it

How deep it goes
Defines the structure and contrasts it with heterogeneous and bipartite graphs, stopping short of detailed algorithmic treatment.
Examples
Uses a recruitment network containing candidate nodes and candidate-relationship edges.
What you can do after
Evaluate whether a network fits a single-node-type, single-edge-type representation before selecting graph methods.
How it connects
Connects homogeneous graphs with heterogeneous graphs, bipartite graphs, graph analysis, and machine-learning methods.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Graph Neural Networks in Action. The section links point to the parts of the book it draws on.