mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Shahid Shaikh <>
Subject Interpret CVB output
Date Fri, 26 Sep 2014 12:36:52 GMT
I have successfully ran CVB JOB with input k (number of topics) as 20 -x
and got the following files as output of the Job

1.      CVB output for parameter “*-o*” which is vector file with 20

2.      “doc-topic-distributions” for parameter “*-dt*” which is a vector
 file with 806 records i.e a record for each input document .

3.      “model-state-after-each-iteration”   for parameter “*-mt*” which
has models after each iteration which has files which is vector file with
20 records.

I need help on how I will relate the documents to topics generated in
output i.e the 20 documents .I believe the 20 topics generated by CVB as
nothing but the end resultant clusters. Please suggest with a approach for
interpreting and mapping the documents into topics/clusters.

Shaikh Shahid G .
+91 9503954781

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