Quick relevance question

[Ryan Gehring <rg...@linkedin.com>]

Hello!
SOLR Newbie here. I managed to pick a use case that is a little goofy for my first SOLR voyage and would love help.

I'd like to rank documents by the total matching term count rather than the usual cosine similarity stuff.

I have a field named "all" which has all my searchable fields copyfielded to it.It seems like I need to use SOLR 4.0 and

_val_:"sum( Termfreq(all, 'term1') , Termfreq(all, 'term2') , … )"

For every query term. Is there a better way to do this ?

Thanks!
Ryan Gehring

Re: Quick relevance question

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

There's actually a Solr JIRA about this:
https://issues.apache.org/jira/browse/SOLR-2953

But it begs the question of why you want to do this?
Are you sure this would actually providing a better
experience for your users? The reason I ask is
that you could put a lot of effort into making this happen
and discover that it was a waste....

IOTW, what's the use case?

Best
Erick

On Sun, Dec 11, 2011 at 5:57 PM, Ryan Gehring <rg...@linkedin.com> wrote:
> Hello!
> SOLR Newbie here. I managed to pick a use case that is a little goofy for my first SOLR voyage and would love help.
>
> I'd like to rank documents by the total matching term count rather than the usual cosine similarity stuff.
>
> I have a field named "all" which has all my searchable fields copyfielded to it.It seems like I need to use SOLR 4.0 and
>
> _val_:"sum( Termfreq(all, 'term1') , Termfreq(all, 'term2') , … )"
>
> For every query term. Is there a better way to do this ?
>
> Thanks!
> Ryan Gehring
>