Tarek Auel created SPARK-7966:
---------------------------------
Summary: add Spreading Activation algorithm to GraphX
Key: SPARK-7966
URL: https://issues.apache.org/jira/browse/SPARK-7966
Project: Spark
Issue Type: Improvement
Components: GraphX
Reporter: Tarek Auel
Priority: Minor
I'm wondering if you would like to add the Spreading Activation algorithm to
GraphX. I have implemented it, using the Pregel-API and would love to share it
with the community.
Spreading activation is a algorithm that was invented to search in associative
networks. The basic idea is, that you have one (or multiple) starting nodes.
The activation spreads out from these nodes to the neighbours and the
neighbours of the neighbours. The activation decreases after every hop. Nodes
that were reached by many activations will have a higher total activation level.
Spreading Activation is for many use cases useful. Imagine you have the social
network of two people. If you apply the spreading activation to this social
graph with the two people as starting nodes, you will get the nodes that are
most important for both.
Some resources:
http://www.websci11.org/fileadmin/websci/posters/105_paper.pdf
https://webfiles.uci.edu/eloftus/CollinsLoftus_PsychReview_75.pdf?uniq=20ou4w
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
