You are viewing a plain text version of this content. The canonical link for it is here.
Posted to solr-user@lucene.apache.org by Jay Potharaju <js...@gmail.com> on 2019/07/08 14:15:35 UTC

Re: Relevance by term position

Thanks, use of payloads works for my use case.
Jay

> On Jun 28, 2019, at 6:46 AM, Alexandre Rafalovitch <ar...@gmail.com> wrote:
> 
> This past thread may be relevant: https://markmail.org/message/aau6bjllkpwcpmro
> It suggests that using SpanFirst of XMLQueryParser will have automatic
> boost for earlier matches.
> The other approach suggested was to use Payloads (which got better
> since the original thread).
> 
> Regards,
>   Alex.
> 
>> On Thu, 27 Jun 2019 at 22:01, Jay Potharaju <js...@gmail.com> wrote:
>> 
>> Hi,
>> I am trying to implement autocomplete feature that should rank documents based on term position in the search field.
>> Example-
>> Doc1- hello world
>> Doc2- blue sky hello
>> Doc3 - John hello
>> 
>> Searching for hello should return
>> Hello world
>> John hello
>> Blue sky hello
>> 
>> I am currently using ngram to do autocomplete. But this does not allow me to rank results based on term position.
>> 
>> Any suggestions on how this can be done?
>> Thanks
>>

Re: Relevance by term position

Posted by Erick Erickson <er...@gmail.com>.
To re-enforce what Alex said, payloads have first-class Solr support as of Solr 6.6, see: https://lucidworks.com/post/solr-payloads/

> On Jul 8, 2019, at 7:15 AM, Jay Potharaju <js...@gmail.com> wrote:
> 
> Thanks, use of payloads works for my use case.
> Jay
> 
>> On Jun 28, 2019, at 6:46 AM, Alexandre Rafalovitch <ar...@gmail.com> wrote:
>> 
>> This past thread may be relevant: https://markmail.org/message/aau6bjllkpwcpmro
>> It suggests that using SpanFirst of XMLQueryParser will have automatic
>> boost for earlier matches.
>> The other approach suggested was to use Payloads (which got better
>> since the original thread).
>> 
>> Regards,
>>  Alex.
>> 
>>> On Thu, 27 Jun 2019 at 22:01, Jay Potharaju <js...@gmail.com> wrote:
>>> 
>>> Hi,
>>> I am trying to implement autocomplete feature that should rank documents based on term position in the search field.
>>> Example-
>>> Doc1- hello world
>>> Doc2- blue sky hello
>>> Doc3 - John hello
>>> 
>>> Searching for hello should return
>>> Hello world
>>> John hello
>>> Blue sky hello
>>> 
>>> I am currently using ngram to do autocomplete. But this does not allow me to rank results based on term position.
>>> 
>>> Any suggestions on how this can be done?
>>> Thanks
>>>