Can the node similarity algorithm calculate the similarity score based on incoming and outgoing relationships from a node?

Neo4j Graph Platform Graph Algorithms/Graph Data Science
1

For example, let's say I have 3 nodes: Dan, Ann and Bob.

Dan LIKES Ann.
Bob LIKES Ann.
Ann LIKES Bob.

I think the algorithm would give Dan and Bob a score of 1 given that they both have outgoing LIKES relationship to Ann. But is there a way to take into consideration the incoming LIKES relationship from Ann to Bob, which would change the similarity score between Dan and Bob?

Thank you

2

Sure - you can load the graph as undirected and then we will compute similarity in both directions. Or you can load LIKES with both natural and reverse orientation (so as LIKES and LIKED_BY) if you want to compare.

Take a peak at our docs on relationship directionality, here: https://neo4j.com/docs/graph-data-science/current/graph-project/#_relationship_orientation