2 Responses to “Atbrox Customer Case Study – Scalable Language Processing with Elastic Mapreduce (Hadoop)”

1. Øystein Nytrø Says:
   November 16th, 2009 at 21:15

   Referring to the customer case, I would guess that n-gram postprocessing and subsequent model-building are the most complex, and depending on n, computationally intensive tasks. Was that
   done locally afterwards, or did you manage to distribute that also?

   — Øystein N.

2. amund Says:
   November 23rd, 2009 at 21:38

   the scope of this project was the creation of n-gram and not the processing afterwards (except some filtering).

Leave a Reply

Name (required)

Mail (will not be published) (required)

Website