You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@mahout.apache.org by shambhusingh <sh...@gmail.com> on 2011/04/21 02:27:46 UTC

TFIDF based on Field

Hi All,

I have a situation where i want to give more relevency to a token from
1 field (say Title) than other column say (Brand).
so is there any way we can manipulate the tfidf value based on the field name.

Thanks in Advance

-- 
Many Thanks,
Shambhu


--
View this message in context: http://lucene.472066.n3.nabble.com/TFIDF-based-on-Field-tp2845023p2845023.html
Sent from the Mahout User List mailing list archive at Nabble.com.

Re: TFIDF based on Field

Posted by Bruce Dou <do...@gmail.com>.
Just add a boost value to the field.

On Thu, Apr 21, 2011 at 8:31 AM, Stanley Xu <we...@gmail.com> wrote:
> repeat the title for multiple times is a simple hack you could do
>
> 在 2011-4-21 上午8:28,"shambhusingh" <sh...@gmail.com>写道:
>> Hi All,
>>
>> I have a situation where i want to give more relevency to a token from
>> 1 field (say Title) than other column say (Brand).
>> so is there any way we can manipulate the tfidf value based on the field
> name.
>>
>> Thanks in Advance
>>
>> --
>> Many Thanks,
>> Shambhu
>>
>>
>> --
>> View this message in context:
> http://lucene.472066.n3.nabble.com/TFIDF-based-on-Field-tp2845023p2845023.html
>> Sent from the Mahout User List mailing list archive at Nabble.com.
>



-- 
A decathlon Drupal developer & programmer
http://blog.eood.cn/

Re: TFIDF based on Field

Posted by Stanley Xu <we...@gmail.com>.
repeat the title for multiple times is a simple hack you could do

在 2011-4-21 上午8:28,"shambhusingh" <sh...@gmail.com>写道:
> Hi All,
>
> I have a situation where i want to give more relevency to a token from
> 1 field (say Title) than other column say (Brand).
> so is there any way we can manipulate the tfidf value based on the field
name.
>
> Thanks in Advance
>
> --
> Many Thanks,
> Shambhu
>
>
> --
> View this message in context:
http://lucene.472066.n3.nabble.com/TFIDF-based-on-Field-tp2845023p2845023.html
> Sent from the Mahout User List mailing list archive at Nabble.com.