mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ted Dunning <>
Subject Re: LSI, cosine and others which use vectors
Date Wed, 24 Jun 2009 16:20:12 GMT
The changes in the weightings are exceedingly small and can be neglected in
practice for quite a long time.

On Wed, Jun 24, 2009 at 6:02 AM, Paul Jones <>wrote:

> >>>> I don't follow how would you do just part of the matrix, i.e if you
> have created the term-doc, surely the weighted data would need to be
> weighted across the whole set, in order to give you relevance, not sure how
> you would slice through a subset of the matrix you want to work with.

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