GraphSAGE node classification training

GraphSAGE is a graph neural network (GNN) architecture that can be used as a supervised algorithm to predict class labels of nodes in a graph. This section provides instructions for how to use the GraphSAGE endpoint for training a model for node classification using Neo4j Graph Analytics for Snowflake.

Syntax

This section covers the syntax used to execute the GraphSAGE node classification training algorithm.

Run GraphSAGE node classification training.
CALL graph.gs_nc_train(
  'CPU_X64_XS',                    (1)
  {
    ['defaultTablePrefix': '...',] (2)
    'project': {...},              (3)
    'compute': {...},              (4)
  }
);
1 Compute pool selector.
2 Optional prefix for table references.
3 Project config.
4 Compute config.
Table 1. Parameters
Name Type Default Optional Description

computePoolSelector

String

n/a

no

The selector for the compute pool on which to run the GraphSAGE node classification training job.

configuration

Map

{}

no

Configuration for graph project, algorithm compute and result write back.

For this algorithm we strongly recommend using a GPU compute pool, unless the dataset is very small and the model shallow.

The configuration map consists of the following three entries.

For more details on below Project configuration, refer to the Project documentation.
Table 2. Project configuration
Name Type

nodeTables

List of node tables.

relationshipTables

Map of relationship types to relationship tables.

Please note that in order for GraphSAGE to properly propagate updates of node embeddings, each type of node must be the target of at least one relationship type. The orientation parameter can be useful to add reverse direction relationships for types of nodes that are only the source of relationships (using the "REVERSE" or "UNDIRECTED" orientations).

Table 3. Compute configuration
Name Type Default Optional Description

target_label

String

n/a

no

The node label (i.e. type) to train to predict on

target_property

String

n/a

no

The node property to train to predict, represented by a column in the input node table of the specified 'target_label'

modelname

String

n/a

no

The name of the model to train (must be unique)

numEpochs

Integer

n/a

no

The number of epochs to train the model

numSamples

List of Integer

n/a

no

The number of neighbors to sample for each layer. Note that this also determines the number of layers

hiddenChannels

Integer

256

yes

The node embedding dimension of the model layers' outputs

activation

String

"relu"

yes

The activation function to use. Valid values are "relu" and "sigmoid"

aggregator

String

"mean"

yes

The neighborhood embedding aggregator to use. Valid values are "mean" and "max"

learningRate

Float

0.001

yes

The learning rate for the optimizer

dropout

Float

0.1

yes

The dropout probability for each layer. Must be a value >= 0.0 and < 1.0

layerNormalization

Boolean

true

yes

Whether to apply layer normalization between the model layers

epochsPerCheckpoint

Integer

max(numEpochs / 10, 1)

yes

The number of epochs between saving model checkpoints

randomSeed

Integer

A random integer

yes

A number used to seed all randomness of the computation

split_ratios

Map

{"TRAIN": 0.6, "TEST": 0.2, "VALID": 0.2}

yes

The ratios as a map to split the target nodes of the input graph into training, test, and validation sets. The keys must be "TRAIN", "TEST" and "VALID". The sum of the values must be 1.0

epochs_per_val

Integer

0

yes

The number of epochs between evaluating the model on the validation set. If set to 0, the model will not be evaluated on the validation set

train_batch_size

Integer

Automatically inferred

yes

The number of target nodes to train on in each batch. If not provided, the algorithm will automatically infer the maximally allowed batch size within the constraints of available memory

eval_batch_size

Integer

train batch size

yes

The batch size to use for evaluation

class_weights

Boolean or Map

false

yes

Whether to use class weights to balance the training data. If set to true, class weights will be calculated based on the distribution of the target labels in the training set. If set to a map, the map must contain the class weight for each target class label

Example

For our example we will use an IMDB dataset with actors, directors, movies, and genres. These all have keywords associated with them, which we will use as features for the nodes. They are connected by relationships where actors act in movies and directors direct movies. The goal is to predict the genre of movies.

We have a database called imdb that contains the tables:

  • actor with columns nodeid and plot_keywords

  • movie with columns nodeid, plot_keywords and genre

  • director with columns nodeid and plot_keywords

  • acted_in with columns sourcenodeid and targetnodeid that represent actor and movie node IDs

  • directed_in with columns sourcenodeid and targetnodeid that represent director and movie node IDs

The plot_keywords columns contain keywords associated with the nodes, encoded as vectors of floats. The genre column contains the target class labels for the movie nodes, which we want to predict.

You can upload this dataset to your snowflake account by following the instructions at github: neo4j-product-examples/snowflake-graph-analytics.

The training query

In the following query we train a GraphSAGE model for node classification on the dataset. We train for 10 epochs, with two hidden layers, and use class weights to balance the class distribution.

To run the query, there is a required setup of grants for the application, your consumer role and your environment. Please see the Getting started page for more on this.

We also assume that the application name is the default Neo4j_Graph_Analytics. If you chose a different app name during installation, please replace it with that.

CALL Neo4j_Graph_Analytics.graph.gs_nc_train('GPU_NV_S', {
    'defaultTablePrefix': 'imdb.gml',
    'project': {
        'nodeTables': ['actor', 'director', 'movie'],
        'relationshipTables': {
            'acted_in': {
                'sourceTable': 'actor',
                'targetTable': 'movie',
                'orientation': 'UNDIRECTED'
            },
            'directed_in': {
                'sourceTable': 'director',
                'targetTable': 'movie',
                'orientation': 'UNDIRECTED'
            }
        }
    },
    'compute': {
        'modelname': 'nc-imdb',
        'numEpochs': 10,
        'numSamples': [20, 20],
        'targetLabel': 'movie',
        'targetProperty': 'genre',
        'classWeights': true
    }
});

The above query should produce a result similar to the one below. The numerical results may vary.

JOB_ID

JOB_START

JOB_END

JOB_RESULT

job_63b8083fc8ef463ab38cd95d2ac345ea

2025-04-29 12:06:28.791

2025-04-29 12:07:10.318

{ "metrics": { "test_acc": 0.7441860437393188, "test_f1_macro": 0.7236689925193787, "test_f1_micro": 0.7441860437393188, "train_acc": 0.9911160469055176, "train_f1_macro": 0.9900508522987366, "train_f1_micro": 0.9911160469055176 } }