[
https://issues.apache.org/jira/browse/LUCENE-5354?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13866989#comment-13866989
]
Michael McCandless commented on LUCENE-5354:
--------------------------------------------
Thanks Remi, the performance seems fine? But I realized this is not the best
benchmark, since all suggestions are just a single token.
New patch looks great; I think we should commit this approach, and performance
improvements can come later if necessary.
bq. see above my comment for your previous suggestion to avoid visiting term
vectors
Oh, the idea I had was to not use term vectors at all: you can get a TermsEnum
for the normal inverted index, and then visit each term from the query, and
then .advance to each doc from the top N results. But we can do this later ...
I'll commit this patch (I'll make some small code style improvements, e.g.
adding { } around all ifs).
> Blended score in AnalyzingInfixSuggester
> ----------------------------------------
>
> Key: LUCENE-5354
> URL: https://issues.apache.org/jira/browse/LUCENE-5354
> Project: Lucene - Core
> Issue Type: Improvement
> Components: modules/spellchecker
> Affects Versions: 4.4
> Reporter: Remi Melisson
> Priority: Minor
> Labels: suggester
> Attachments: LUCENE-5354.patch, LUCENE-5354_2.patch,
> LUCENE-5354_3.patch
>
>
> I'm working on a custom suggester derived from the AnalyzingInfix. I require
> what is called a "blended score" (//TODO ln.399 in AnalyzingInfixSuggester)
> to transform the suggestion weights depending on the position of the searched
> term(s) in the text.
> Right now, I'm using an easy solution :
> If I want 10 suggestions, then I search against the current ordered index for
> the 100 first results and transform the weight :
> bq. a) by using the term position in the text (found with TermVector and
> DocsAndPositionsEnum)
> or
> bq. b) by multiplying the weight by the score of a SpanQuery that I add when
> searching
> and return the updated 10 most weighted suggestions.
> Since we usually don't need to suggest so many things, the bigger search +
> rescoring overhead is not so significant but I agree that this is not the
> most elegant solution.
> We could include this factor (here the position of the term) directly into
> the index.
> So, I can contribute to this if you think it's worth adding it.
> Do you think I should tweak AnalyzingInfixSuggester, subclass it or create a
> dedicated class ?
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
