[ 
https://issues.apache.org/jira/browse/MAHOUT-939?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13182494#comment-13182494
 ] 

Grant Ingersoll commented on MAHOUT-939:
----------------------------------------

bq. Are these results with held-out data? Or are they what is reported by the 
cross fold learners?

Here are the steps we do currently:
# Convert mail archives to sequence files
# Encode the text using seq2encoded: {code}$MAHOUT seq2encoded --input 
$MAIL_OUT --output $SEQ2SP --analyzerName 
org.apache.mahout.text.MailArchivesClusteringAnalyzer --cardinality 100000{code}
# Do some minor reworking of the vectors to get decent labels
# Split the vectors into a train and test set, randomizing as we go thanks to 
MAHOUT-904 {code}$MAHOUT split --input $SEQ2SPLABEL --mapRedOutputDir 
$MAPREDOUT  --randomSelectionPct 20 --overwrite --sequenceFiles --method 
mapreduce{code}
# Run the training
# Run the test

All of this is in examples/bin/asf-email-examples.sh

The code for TrainASFModel more or less mirrors what is in the 20NewsGroup by 
using ALR.  I even refactored the two a bit to share some common code.  
Hopefully I didn't mess things up there despite the tests passing and the 
results looking reasonable.
                
> ASF Email Classification Examples don't always produce good results
> -------------------------------------------------------------------
>
>                 Key: MAHOUT-939
>                 URL: https://issues.apache.org/jira/browse/MAHOUT-939
>             Project: Mahout
>          Issue Type: Bug
>    Affects Versions: 0.6
>            Reporter: Grant Ingersoll
>            Assignee: Grant Ingersoll
>              Labels: MAHOUT_INTRO_CONTRIBUTE
>             Fix For: 0.7
>
>         Attachments: 939.patch, MAHOUT-939.patch, MAHOUT-939.patch, 
> MAHOUT-939.patch, strip_reject.patch
>
>
> The classification examples for the ASF email don't work all that well 
> currently in terms of quality when it comes to more than a few labels.  Also, 
> need to determine how much memory is required for vectors of cardinality size 
> 100K.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Reply via email to