I have a question about the scope_train and scope_test options that come
with discriminate.pl.

Now, my impression is that scope_train defines a window around the target
word from which features can be selected. And that scope_test defines a
window around a test instance that indicates which words will contribute
vectors to the averaged vector that represents the context (in order 2 at
least). In the case of order 1, scope_test indicate a window of words in
which features are identified.

Is all of that accurate?

If so, on to the question. What if you are using a single data file as
both your test and training data. In other words, you are extracting
features from the data you are clustering. What is the effect of
scope_train and scope_test in that situation? Is it the same, in that
scope_train defines the window in which features are extracted, and
scope_test defines the window in which features are matched/represented
using either the order 1 or order 2 style.

For a while I was confused, and thought that scope_train would have no
effect when you identified features from the data you were clustering,
but I've seen results that suggest that this isn't the case.

Thanks!
Ted

--
Ted Pedersen
http://www.d.umn.edu/~tpederse


-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
_______________________________________________
senseclusters-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/senseclusters-users

Reply via email to