change documents' label

[jiuren]

hi,all

I'm doing a multilabel classification and I split the task to some
small tasks. I need to modify the labels of some documents in one
corpus to build a new corpus. As you know, building a corpus from raw
files is time-consuming, especially for CJK languages which need to be
segmented first.

But there is no direct api to do this work. I've read part of the
source code and found a setLabel() method which is not public. Can I
just modify the method to public and let the user code use it?Will
this bring some side effect?

another question:
If I use multiFieldDocument, can I make use some fields only to store
some meta data but not to affect the classification? For example I can
set some field unindexed in lucene.

[DigitalPebble]

But there is no direct api to do this work. I've read part of the
source code and found a setLabel() method which is not public. Can I
just modify the method to public and let the user code use it?Will
this bring some side effect?

I don't expect so. Why don't you try?
 

another question:
If I use multiFieldDocument, can I make use some fields only to store
some meta data but not to affect the classification? For example I can
set some field unindexed in lucene.

I suppose we could define a special type of field with would be ignored when generating the vector. Feel free to file an issue in github and send a patch

J.

--

 
Open Source Solutions for Text Engineering
 
http://digitalpebble.blogspot.com
http://www.digitalpebble.com