Introduction documentation feedback

[GoogleCodeExporter]

I received the following feedback from a developer new to UIMA, NLP, and
machine learning - though otherwise very sharp.  I think it would be
worthwhile to address all of the points that he makes.

<begin-message>
I looked at the wiki earlier and was a bit overwhelmed
and abandoned it as my first source of information.
After a short conversation with Kevin, it makes a bit
more sense, but I think it would benefit from a paragraph
or two of introduction. I may be on the edge of the expected
audience, but following are some questions I have after looking
at the main page and the main wiki page. A better introduction
on googleCode may not answer them all.

BTW, Do we have a book in the lab library that introduces machine
learning in the context of NLP? I've read Jackson and Moulinier.

http://code.google.com/p/cleartk/

"...feature extraction library" ...like what? POS, named entity, 
misc relationships?

" ...wrappers..." *UIMA* wrappers?
I'd like to learn more about maximum entropy, support vector machines
and conditional random fields, but wouldn't expect that from a ClearTK
intro.

...also sequential taggers, chunkers, role labelling and temporal
resolution.

- Where does the name come from? (certainly not Tcl/Tk)

 http://code.google.com/p/cleartk/w/list

- What's a classifier? ...I'm guessing you could use one
to do tagging in UIMA.

-What's the Maxent classifier and how is it different
than the POS tagger?

-What's a chunk tokenizer and how is that different from
other kinds of tokenizers.

- How is ClearTK both a pos tagger and these other things?

Original issue reported on code.google.com by pvogren@gmail.com on 30 Jan 2009 at 10:06

[GoogleCodeExporter]

Original comment by steven.b...@gmail.com on 6 Feb 2009 at 6:53

• Added labels: Priority-Critical
• Removed labels: Priority-Medium

[GoogleCodeExporter]

I have not gone through this message carefully to make sure each point is 
thoroughly
addressed.  However, I think the ConceptualOverview wiki page does a reasonably 
good
job of addressing the main issues raised here.  Please see the wiki.

Original comment by pvogren@gmail.com on 7 Feb 2009 at 3:46

• Changed state: Fixed