lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Grant Ingersoll <gsing...@apache.org>
Subject Re: Solr Queries
Date Thu, 12 Nov 2009 16:05:56 GMT

On Nov 12, 2009, at 8:55 AM, Rakhi Khatwani wrote:

> Hi,
>         I am using solr 1.3 and i hv inserted some data in my comment
> field.
> for example:
> 
> for document1:
> <str name="comment">
> The iPhone 3GS finally adds common cell phone features like multimedia
> messaging, video recording, and voice dialing. It runs faster; its promised
> battery life is longer; and the multimedia quality continues to shine.
> 
> 
> The iPhone 3GS' call quality shows no improvements and the 3G signal
> reception remains uneven. We still don't get Flash Lite, USB transfer and
> storage, or multitasking.
> </str>
> 
> 
> for document2:
> <str name="comment">
> Sony Ericsson c510 has 3.2MP cybershot camera with smile detectino. Amazing
> phone, faster than Sony Ericsson w580iSony Ericcsion w580i camera is only
> 2MP with no autofocus and smile detection. it doesnot even have a flash
> leading to poor quality pictures
> </str>
> 
> A]
> 
> now when i apply the following queries, i get 0 hits:
> 1.comment:iph*e
> 2.comment:iph?ne

What field type are you using?  This is in your schema.xml

> 
> B] Can i apply range queries on part of the content?

> 
> C] Can i apply more the one wildcard in a query?? for example comment:ip*h*
> (this command works but its equivalent to just using 1ipho*)

Yes.

> 
> D] for fuzzy queries:
>     content:iphone~0.7 returns both the documents.
>     content:iphone~0.8 returns no documents (similarly for 0.9).
> 

The fuzz factor there incorporates the edit distance.  I gather the first Sony doc has a match
on phone and the score is between 0.7 and 0.8.  You can add &debugQuery=true to see the
explains. 

> However if i change it to iPhone,
>    content:iPhone~0.7 returns 0 documents
>    content:iPhone~0.5 returns both the documents.
> 
> Is fuzzy search case sensitive? even if it is, why am i not able to retrieve
> unexpected results.

Again, this all comes back to how you analyze the documents based on what Field Type you are
using?

--------------------------
Grant Ingersoll
http://www.lucidimagination.com/

Search the Lucene ecosystem (Lucene/Solr/Nutch/Mahout/Tika/Droids) using Solr/Lucene:
http://www.lucidimagination.com/search


Mime
View raw message