[
https://issues.apache.org/jira/browse/LUCENE-3980?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13253558#comment-13253558
]
Ian Pooley commented on LUCENE-3980:
------------------------------------
The problem is that the queries are generated by a QueryParser from large,
complex query strings created by our internal users. Most of these queries
return exactly want they expect but, a couple of days ago, one of the users
noticed that "A B"~5 within one of these queries returned slightly different
results from a query that was identical other than the clause that as "B A"~5.
Now that I have a better idea as to what is going on under the covers, my
challenge is to translate that into non-technical rules that will allow all
permutations of "A B C D E..."~n to give the desired answer.
> Word order seems to affect proximity searching
> ----------------------------------------------
>
> Key: LUCENE-3980
> URL: https://issues.apache.org/jira/browse/LUCENE-3980
> Project: Lucene - Java
> Issue Type: Bug
> Components: core/search
> Reporter: Ian Pooley
> Priority: Minor
>
> It would appear that the order of words within a search query affects a
> proximity search.
> For instance, for the text "The proximity operator seems to match differently
> based on word order", a match is found for "proximity order"~8 but is not
> found for "order proximity"~8. In order for the latter to find a match, it
> needs to be changed to "order proximity"~10.
> Both the text and the query are processed using
> org.apache.lucene.analysis.standard.StandardAnalyzer.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
