GraphSAGE
Glossary
- Directed
-
Directed trait. The algorithm is well-defined on a directed graph.
- Undirected
-
Undirected trait. The algorithm is well-defined on an undirected graph.
- Heterogeneous nodes
-
Heterogeneous nodes fully supported. The algorithm has the ability to distinguish between nodes of different types.
- Heterogeneous relationships
-
Heterogeneous relationships fully supported. The algorithm has the ability to distinguish between relationships of different types.
- Weighted relationships
-
Weighted trait. The algorithm supports a relationship property to be used as weight, specified via the relationshipWeightProperty configuration parameter.
- Node properties
-
Node properties trait. The algorithm makes use of node properties.
- CPU
-
The algorithm runs on CPU compute pools.
- GPU
-
The algorithm runs on GPU compute pools.
Introduction
GraphSAGE is a graph neural network (GNN) architecture outlined in the paper "Inductive Representation Learning on Large Graphs" by Hamilton et al. (2017). In Neo4j Graph Analytics for Snowflake, GraphSAGE is exposed as algorithms for both node classification and unsupervised for generating node embeddings.
GraphSAGE for node classification
The GraphSAGE architecture can be used as a supervised algorithm to predict labels of nodes in a graph. This subchapter provides instructions for how to use the GraphSAGE endpoints for node classification:
-
GraphSAGE node classification training — Training a GraphSAGE model for node classification
-
GraphSAGE node classification prediction — Using a trained GraphSAGE model for node classification to make new predictions
GraphSAGE for node embeddings
The GraphSAGE architecture can be used as an unsupervised algorithm to generate embeddings for nodes in a graph. This subchapter provides instructions for how to use the GraphSAGE node embeddings endpoints:
-
GraphSAGE node embedding training — Training a GraphSAGE model for node embeddings
-
GraphSAGE node embedding prediction — Using a trained GraphSAGE model for node embeddings to infer new embeddings
Moreover, please refer to Model catalog for how to show, describe and drop trained models.