Neo4j GDS Java API

Neo4j Graph Platform
1

Hello everyone,
i am new to graph databases. Recently i became a bit familiar with the cypher language, i played with a few datasets in order to practice crud operations with neo4j. I would like to create a java project that predicts links between nodes (i'm gonna pick a small dataset at first). I have read FastRP and kNN example which is quite simple. My question is, is there any way that i can use these algorithms from a java library? I was searching around for tutorial articles or youtube videos but didn't find anything useful. I also checked for Maven Repositories, found a lot of them, but there is no documentation of how to use them.
I am using neo4j community version 5.13.0 combined with neo4j-graph-data-science-2.5.2.
Thanks in advance.

2

I think your option is to use the java driver.

3

Well right now this was my way to go, connect to a neo4j driver that has GDS plugin installed and execute queries like:

CALL gds.graph.project(
  'purchases',
  ['Person','Product'],
  {
    BUYS: {
      orientation: 'UNDIRECTED',
      properties: 'amount'
    }
  }
)

CALL gds.fastRP.mutate('purchases',
  {
    embeddingDimension: 4,
    randomSeed: 42,
    mutateProperty: 'embedding',
    relationshipWeightProperty: 'amount',
    iterationWeights: [0.8, 1, 1, 1]
  }
)
YIELD nodePropertiesWritten

(code quotes are from FastRP and kNN example). I was looking for a high level java way. I can see that there is a Python Client, maybe there is a similar library in java?

4

Yes, neo4j added that GDS support to only the python driver. Probably because they think data scientists use python.