Vector encoding and classifying of Text

[Sarath P R <sa...@gmail.com>]

Hi all,

I have a doubt. Is it enough if i use a particular vector encoder during
training a model and  use the same encoding for classifying a text document
against the model. I am not able to get correct class label.

I used seq2encoded instead of seq2sparse during training. is it right to do
so ? like this

mahout-distribution-0.7/bin/mahout seqdirectory -i /home/sarath/MyData -o
/home/sarath/MyModel-seq
mahout-distribution-0.7/bin/mahout seq2encoded -i /home/sarath/MyModel-seq
-o /home/sarath/MyModel-vectors -ec
org.apache.mahout.vectorizer.encoders.StaticWordValueEncoder
mahout-distribution-0.7/bin/mahout trainnb -i /home/sarath/MyModel-vectors
-el -o /home/sarath/Model-1 -li /home/sarath/labelindex-1

Thanks in advance

-- 
Thank You
Sarath P R | cell +91 99 95 02 4287 | http://sprism.blogspot.com