lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Erik Hatcher <erik.hatc...@gmail.com>
Subject Re: Question about relevance
Date Fri, 08 Jan 2010 11:02:36 GMT
One technique I've seen commonly used is to index both stemmed and  
unstemmed fields, and during search query both and boost the unstemmed  
field matches higher.

	Erik

On Jan 8, 2010, at 4:05 AM, Yannick Caillaux wrote:

> Hi,
>
> I index 2 documents. the first contains the word "Wallis" in the  
> title field. The second has the same title but "Wallis" is replaced  
> by "Wall".
> I execute the query : "title:wallis"
> During the search, "Wallis" is cut by the FrenchAnalyzer and becomes  
> "wall". So the two documents are results for the search.
>
> My problem is : the two results have the same relevance.
> I thought that the document containing "Wallis" would have better  
> relevance because I search for the word "wallis" and not "wall".
>
> Relevance is calculated from the searched word (wallis) or from the  
> analyzed word (wall)? Is there any solution to get better relevance  
> for the result "wallis" ?
> For information i'm in lucene 2.3.2.
>
> Thanks
>
> Yannick
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org
>


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


Mime
View raw message