If opinion leader following a lot of accounts probability that he get transmitted information decrease. So i'm wondering does this is taken into account in Eigenvector graph centrality algorithm?
Welcome to the Neo4j community!
When you use the eigenvector algorithm, you specify incoming, outgoing, or both. The value produced is based only on the direction specified. So if you are analyzing incoming relationships, outgoing relationships are not accounted for. If you specify both, then all relationships are used for the value returned, but there is no differentiation between incoming or outgoing.
So to answer your question, the engenvalue result for a node that has followers is not adjusted if that node follows others.
Thank you, Elaine.
What if i will set weight so nodes with large amount of outgoin nodes have lower score than nodes that have small amount of outgoing nodes.
In what cases i can use weights ?
May i set weight that is equal to avarage posts reaction count of influencer?
Yes setting the weights to be lower for outgoing and higher for incoming would do it provided you use both for the direction of the analysis.