ahhh I see..good point..yes, for a high number of unique scores the
secondary sort won't have any effect..

On 30 November 2010 09:32, Jason Brown <jason.br...@sjp.co.uk> wrote:

> Hi - you do understand may case - we tried what you suggested but as the
> relevancy is very precise we couldn't get it it to do a dual-sort.
>
> I like the idea of using one of the dismax parameters (bf) to in-effect
> increase the boost on a newer document.
>
> Thanks for all replies, most useful.
>
>
> -----Original Message-----
> From: Savvas-Andreas Moysidis [mailto:
> savvas.andreas.moysi...@googlemail.com]
> Sent: Tue 30/11/2010 09:26
> To: solr-user@lucene.apache.org
> Subject: Re: Boost on newer documents
>
> hi,
>
> I might not understand your case right but can you not add an extra
> publishedDate field and then specify a secondary (after relevance) sort by
> that?
>
> On 30 November 2010 08:05, <jan.kure...@nokia.com> wrote:
>
> > You could also put a short representation of the data (I suggest days
> since
> > 01.01.2010) as payload and calculate boost with payload function of the
> > similarity.
> >
> > >-----Original Message-----
> > >From: ext Jason Brown [mailto:jason.br...@sjp.co.uk]
> > >Sent: Montag, 29. November 2010 17:28
> > >To: solr-user@lucene.apache.org
> > >Subject: Boost on newer documents
> > >
> > >
> > >Hi,
> > >
> > >I use the dismax query to search across several fields.
> > >
> > >I find I have a lot of documents with the same document name (one of the
> > fields that the dismax queries) so I wanted to adjust the
> > >relevance so that titles with a newer published date have a higher
> > relevance than documents with the same title but are older. Does
> > >anyone know how I can achieve this?
> > >
> > >Thank You
> > >
> > >Jason.
> > >
> > >If you wish to view the St. James's Place email disclaimer, please use
> the
> > link below
> > >
> > >http://www.sjp.co.uk/portal/internet/SJPemaildisclaimer
> >
>
>
> If you wish to view the St. James's Place email disclaimer, please use the
> link below
>
> http://www.sjp.co.uk/portal/internet/SJPemaildisclaimer
>

Reply via email to