You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@mahout.apache.org by Taner Diler <ta...@gmail.com> on 2013/07/26 16:58:25 UTC

About to find category of a new article

Hi all,

I want to be sure about a subject.

I've lots of articles about sports, mobile technologies, beverage & food,
automotive...

When I take a new article, system should tell me that this is about
beverage & food.

Classification is doing this, am I right? Is there a sample or tutorial
about like kind of this usage?

Re: About to find category of a new article

Posted by Ted Dunning <te...@gmail.com>.
The basic problem is, indeed, classification.

There are many ways to generate the predictor variables used by the
classifier.

LSA is one such way to do this.

In mahout, you can try naive bayes or logistic regression.



On Fri, Jul 26, 2013 at 8:02 AM, Peyman Mohajerian <mo...@gmail.com>wrote:

> LSA does that, http://en.wikipedia.org/wiki/Latent_semantic_analysis, you
> can SSVD in mahout to do it.
>
>
> On Fri, Jul 26, 2013 at 7:58 AM, Taner Diler <ta...@gmail.com>
> wrote:
>
> > Hi all,
> >
> > I want to be sure about a subject.
> >
> > I've lots of articles about sports, mobile technologies, beverage & food,
> > automotive...
> >
> > When I take a new article, system should tell me that this is about
> > beverage & food.
> >
> > Classification is doing this, am I right? Is there a sample or tutorial
> > about like kind of this usage?
> >
>

Re: About to find category of a new article

Posted by Peyman Mohajerian <mo...@gmail.com>.
LSA does that, http://en.wikipedia.org/wiki/Latent_semantic_analysis, you
can SSVD in mahout to do it.


On Fri, Jul 26, 2013 at 7:58 AM, Taner Diler <ta...@gmail.com> wrote:

> Hi all,
>
> I want to be sure about a subject.
>
> I've lots of articles about sports, mobile technologies, beverage & food,
> automotive...
>
> When I take a new article, system should tell me that this is about
> beverage & food.
>
> Classification is doing this, am I right? Is there a sample or tutorial
> about like kind of this usage?
>