[
https://issues.apache.org/jira/browse/SOLR-14590?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17144109#comment-17144109
]
Tomas Eduardo Fernandez Lobbe commented on SOLR-14590:
------------------------------------------------------
Discussed this briefly with [~varun] yesterday via Slack. For queries, the idea
is to have a new query parser that can take the parameters and function to use,
and can be added either to a {{bq}} parameter in a {{dismax}} query, or can be
added to a {{lucene}} query using subqueries, like using {{_query_}} magic
field. Something like
{code:java}
bq={!rank f=myFeatureField function=log weight=2 scalingFactor=1}pagerank
{code}
For index time, we could have a syntax like {{featureName:featureValue}}. So,
when someone is using SolrJ they would just add a field like:
{code:java}
doc.setField("myFeatureField", "pagerank:2");
{code}
> Add support for FeatureField in Solr
> ------------------------------------
>
> Key: SOLR-14590
> URL: https://issues.apache.org/jira/browse/SOLR-14590
> Project: Solr
> Issue Type: Improvement
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Tomas Eduardo Fernandez Lobbe
> Priority: Major
>
> {{FeatureField}}’s allow users to store scoring factors in the index that are
> encoded as term frequencies. By doing this, {{FeatureFields}} can be used in
> combination with the {{minExactCount}} parameter to skip over non-competitive
> documents, and produce results faster. See LUCENE-8197
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
