[
https://issues.apache.org/jira/browse/LUCENE-4100?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16203693#comment-16203693
]
Robert Muir commented on LUCENE-4100:
-------------------------------------
yeah, I think i am looking at it from the top-down (indexsearcher) vs bottom up
(queries).
indexsearcher already knows if scores are needed (e.g. from the Sort), but
there is no way to tell it that approximate total hit count is acceptable. If
we can do that, then I think we can make the early termination case really easy
for the sorted case, index order case, and also this maxscore case.
{quote}
Ideally we would not need another parameter on Query.createWeight for MAXSCORE
either, but the issue is that depending on whether you want to collect all hits
or only the top-scoring ones, then we need different Scorer impls.
{quote}
we do? (genuine question). We added needsScores because previously scorers had
to always be ready for you to "lazily" call score(), and this prevented scoring
from doing much more interesting things up-front like caching whole bitsets,
but is it really the case for maxScore? I'm just asking because the new scorer
here looks a hell of a lot like a disjunction scorer :) If we truly need a
different impl, we should maybe still think it thru because of stuff like
setMinCompetitiveScore() method, which would make no sense except for that
case. I do like that in your patch AssertingScorer checks all that stuff, but
there it is a bit confusing.
> Maxscore - Efficient Scoring
> ----------------------------
>
> Key: LUCENE-4100
> URL: https://issues.apache.org/jira/browse/LUCENE-4100
> Project: Lucene - Core
> Issue Type: Improvement
> Components: core/codecs, core/query/scoring, core/search
> Affects Versions: 4.0-ALPHA
> Reporter: Stefan Pohl
> Labels: api-change, gsoc2014, patch, performance
> Fix For: 4.9, 6.0
>
> Attachments: LUCENE-4100.patch, LUCENE-4100.patch,
> contrib_maxscore.tgz, maxscore.patch
>
>
> At Berlin Buzzwords 2012, I will be presenting 'maxscore', an efficient
> algorithm first published in the IR domain in 1995 by H. Turtle & J. Flood,
> that I find deserves more attention among Lucene users (and developers).
> I implemented a proof of concept and did some performance measurements with
> example queries and lucenebench, the package of Mike McCandless, resulting in
> very significant speedups.
> This ticket is to get started the discussion on including the implementation
> into Lucene's codebase. Because the technique requires awareness about it
> from the Lucene user/developer, it seems best to become a contrib/module
> package so that it consciously can be chosen to be used.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
