Amazon like search function

[nicksnels <ni...@gmail.com>]

Hi,

I'm looking for ideas on how to implement an Amazon.com like search
function. If for example you do a search on - optimization algorithm java.
Amazon first returns results that contain all of these keywords, then
followed by a list that contains both the words "optimization algorithm",
then a list with books that contain the words "algorithm java" and finally a
list with books that contain both the words "optimization java". 

At the moment I'm using the solrQueryParser defaultOperator="OR" option, but
this returns to many results. I have tried to switch to the "AND" option but
that option returns to few results. So I would like to present the results
like Amazon does, so the user has some more options to choose from.

Kind regards,

Nick
-- 
View this message in context: http://lucene.472066.n3.nabble.com/Amazon-like-search-function-tp1940223p1940223.html
Sent from the Solr - User mailing list archive at Nabble.com.

Re: Amazon like search function

[Otis Gospodnetic <ot...@yahoo.com>]

Hi Nick,

You can also have a look at 
http://sematext.com/products/dym-researcher/index.html , which will help.
That particular piece doesn't do *exactly* what you are describing, but it also 
includes a component that does pretty much what you are after (even though it is 
not described at the above URL) - we call it query relaxing.

Otis
----
Sematext :: http://sematext.com/ :: Solr - Lucene - Nutch
Lucene ecosystem search :: http://search-lucene.com/



----- Original Message ----
> From: nicksnels <ni...@gmail.com>
> To: solr-user@lucene.apache.org
> Sent: Sun, November 21, 2010 9:08:52 AM
> Subject: Amazon like search function
> 
> 
> Hi,
> 
> I'm looking for ideas on how to implement an Amazon.com like  search
> function. If for example you do a search on - optimization algorithm  java.
> Amazon first returns results that contain all of these keywords,  then
> followed by a list that contains both the words "optimization  algorithm",
> then a list with books that contain the words "algorithm java"  and finally a
> list with books that contain both the words "optimization  java". 
> 
> At the moment I'm using the solrQueryParser defaultOperator="OR"  option, but
> this returns to many results. I have tried to switch to the "AND"  option but
> that option returns to few results. So I would like to present the  results
> like Amazon does, so the user has some more options to choose  from.
> 
> Kind regards,
> 
> Nick
> -- 
> View this message in context: 
>http://lucene.472066.n3.nabble.com/Amazon-like-search-function-tp1940223p1940223.html
>
> Sent  from the Solr - User mailing list archive at Nabble.com.
> 

Re: Amazon like search function

[Erick Erickson <er...@gmail.com>]

Take a look at DisMax, particularly the mm parameter
(minimum should match). It's kinda confusing, but I
it might help...

See: http://wiki.apache.org/solr/DisMaxQParserPlugin

I believe 3.x as well as trunk also have "Extended Dismax"
or eDismax that provides some additional capabilities.

Best
Erick

On Sun, Nov 21, 2010 at 9:08 AM, nicksnels <ni...@gmail.com> wrote:

>
> Hi,
>
> I'm looking for ideas on how to implement an Amazon.com like search
> function. If for example you do a search on - optimization algorithm java.
> Amazon first returns results that contain all of these keywords, then
> followed by a list that contains both the words "optimization algorithm",
> then a list with books that contain the words "algorithm java" and finally
> a
> list with books that contain both the words "optimization java".
>
> At the moment I'm using the solrQueryParser defaultOperator="OR" option,
> but
> this returns to many results. I have tried to switch to the "AND" option
> but
> that option returns to few results. So I would like to present the results
> like Amazon does, so the user has some more options to choose from.
>
> Kind regards,
>
> Nick
> --
> View this message in context:
> http://lucene.472066.n3.nabble.com/Amazon-like-search-function-tp1940223p1940223.html
> Sent from the Solr - User mailing list archive at Nabble.com.
>