mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "petar.mitrovic" <>
Subject Re: How to determine which cluster an item belongs to
Date Tue, 27 Dec 2011 14:40:31 GMT
Hi Abin,

Thank you for your reply.

As I mentioned, I have already tried to iterate over the sequence file as
you suggested. But this does not solve my essential problem which is how to
get article ids from the same cluster (not vectors). Or at least I can not
see how can I use it.

Maybe I can extract non-zero elements from the vector data and create Lucene
query to find appropriate article. But this seems like big overhead to me.
Instead of that, I am thinking of implementing my own cluster writer which
will store pairs like {clusterId, articleId} instead of {clusterId, Vector}.

Any clarification on your idea would be appreciated, as well as any other


View this message in context:
Sent from the Mahout User List mailing list archive at

View raw message