mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Juan José Ramos <jjar...@gmail.com>
Subject Re: Wiki - 'Quick tour of text analysis using the Mahout command line' clarification
Date Tue, 25 Feb 2014 15:24:18 GMT
Cool. Thanks for the clarification.


On Tue, Feb 25, 2014 at 3:18 PM, Suneel Marthi <suneel_marthi@yahoo.com>wrote:

> That's a mistake on wiki that needs to be corrected. U r tight it should
> be the similarity.
>
> Each row would have the 10 most similar docs  for ever doc.
>
>
>
> Sent from my iPhone
>
> > On Feb 25, 2014, at 9:22 AM, Juan José Ramos <jjarmos@gmail.com> wrote:
> >
> > In the wiki page: 'Quick tour of text analysis using the Mahout command
> > line'.
> >
> >
> https://cwiki.apache.org/confluence/display/MAHOUT/Quick+tour+of+text+analysis+using+the+Mahout+command+line
> >
> > At the very bottom it is said that
> >
> >   1. This will generate the 10 most similar docs to each doc in the
> >   collection.
> >
> >
> >   1. Examine the similarity list:
> >   mahout seqdumper -i reuters-matrix/matrix | more
> >
> >
> > Instead of reuters-matrix/matrix, shouldn't it be reuters-similarity/
> > part-r-00000 since that is the file of the output of rowsimilarity? Or
> does
> > on the contrary the rowsimilarity tool also write to reuters-matrix/?
> >
> > I would expect to contain the 10 most similar documents for every
> document
> > in the reuters' catalogue. Is that correct?
> >
> > Many thanks.
> > Juanjo.
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message