Showing results for 
Search instead for 
Did you mean: 

Graph Data Science: Filtered Node Similarity

Node Clone

Is there a Filtered Node Similarity Example example that i can reference? I am having difficulty configuring node filters for a bipartite projection. From what i can find, Neo4j current documentation does not include an example and truncates some of the syntax for TargetNodeFilters and SourceNodeFilters. For one of my labels, I have projected a numerical vector that resulted from one-hot encoding. For this label, i want to filter the nodes as input to the similarity algorithm using a node filter on that numerical vector. For example:

MATCH (p:Person) where p.vector[0] = 1 AS people
sourceNodeFilter: people

In the node similarity call, where do the above cypher "fit" in the call? Here is what is currently not working...


{ degreeCutoff: 30,
relationshipWeightProperty: 'count',
'MATCH (p:Person) WHERE p.vector[0] = 1 as people',
sourceNodeFilter: 'people'
YIELD node1, node2, similarity


Node Clone

I have changed the similarity call
from: CALL
to: CALL
and still not sure if call is working correctly

Additonal question about how to use "sourceNodeFilter"

    Relating this topic, I am struggling with the appropriate use of Cypher relating to Filtered Node Similarity Algo.
    I would appreciate if someone would teach me how to use “sourceNodeFilter” (or “targetNodeFileter” ) of Filtered Node Similarity algo.

      I want to use the following option for “sourceNodeFilter” (or “targetNodeFilter”)

a list of nodes

MATCH (person:Person) WHERE person.age > 35 collect(person) AS people … sourceNodeFilter: people

        However I do not understand how I should exactly write the above in the query sentence. Could someone give me an example?
        My query which is not working is like this:

{ MATCH (person:Person) WHERE person.age = 35 collect(person) AS people, sourceNodeFilter:people,

YIELD node1, node2, similarity


Nodes 2022
NODES 2022, Neo4j Online Education Summit

On November 16 and 17 for 24 hours across all timezones, you’ll learn about best practices for beginners and experts alike.