You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "sachin (JIRA)" <ji...@apache.org> on 2016/11/28 19:49:58 UTC

[jira] [Created] (SOLR-9806) page wise probabilistic elevation of result documents based on combination of signal values

sachin created SOLR-9806:
----------------------------

             Summary: page wise probabilistic elevation of result documents based on combination of signal values
                 Key: SOLR-9806
                 URL: https://issues.apache.org/jira/browse/SOLR-9806
             Project: Solr
          Issue Type: Bug
      Security Level: Public (Default Security Level. Issues are Public)
            Reporter: sachin


We had a requirement to elevate result documents based on individual signal values and after searching thoroughly into available Solr modules, we couldn't find anything and therefore decided to write our own module.

Problem Description: We wanted to boost results in a probabilistic fashion. i.e x% of the results on each page must have a signal value above threshold value. One immediate usecase that can be thought of is recent results. We want to boost results which are above a certain threshold such that x% of the results in each page belong to these recent documents.

Solution: We extend the solrQueryElevationComponent to support our probabilistic elevation component. An additional comparator is added and the overall runTime changes from nlogn to 2nlogn



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org