You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Vasia Kalavri (JIRA)" <ji...@apache.org> on 2016/04/19 19:07:25 UTC
[jira] [Commented] (FLINK-3780) Jaccard Similarity
[ https://issues.apache.org/jira/browse/FLINK-3780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15248193#comment-15248193 ]
Vasia Kalavri commented on FLINK-3780:
--------------------------------------
Hi [~greghogan],
what do you mean by "all non-zero" similarity scores? Is the objective here to compute the jaccard similarity of all vertices against all other vertices in the input graph?
> Jaccard Similarity
> ------------------
>
> Key: FLINK-3780
> URL: https://issues.apache.org/jira/browse/FLINK-3780
> Project: Flink
> Issue Type: New Feature
> Components: Gelly
> Affects Versions: 1.1.0
> Reporter: Greg Hogan
> Assignee: Greg Hogan
>
> Implement a Jaccard Similarity algorithm computing all non-zero similarity scores. This algorithm is similar to {{TriangleListing}} but instead of joining two-paths against an edge list we count two-paths.
> {{flink-gelly-examples}} currently has {{JaccardSimilarityMeasure}} which relies on {{Graph.getTriplets()}} so only computes similarity scores for neighbors but not neighbors-of-neighbors.
> This algorithm is easily modified for other similarity scores such as Adamic-Adar similarity where the sum of endpoint degrees is replaced by the degree of the middle vertex.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)