[agtk-devel] Re: Question about events and format.

[Haejoong L. <hae...@un...>]

Dear Fredrik,

I'm copying this message to agtk developer mailing list.  I have
an answer for the first question.  For the second question, I'd
like to hear other senior members' opinions.  By the way, I don't
have a Ph.D. degree.  (I don't know if I want to have one :-)

There are two methods we have been using to label a time point.
In the first method, the time point label becomes a feature of
other annotations.  For example, suppose an event, say event_01,
which occurs at the boundary of words.  In the following
annotation graph, to label event_01 that occurs at 14.1 sec can
be represented as follows:

     ____   ____   ____   ____
    / 22 \ / 23 \ / 24 \ / 25 \   ---> 'word' type annotations
   o      o      o      o      o
  12.7   13.2   13.9   14.1   15.0

  feature of annotation 24:  text = "cat"
                             pos = "noun"
                             event_01 = "true"
                             ...


In the second method, we make an annotation, of which the start
and end anchors have the same time.  In this method, the
assumption that event_01 orrurs only at the boundary of words
is not necessary.  Here is the example:

                    ______________
     ____   ____   / 24  ___      \
    / 22 \ / 23 \ /     /25 \      \
   o      o      o     o     o      o
  12.7   13.2   13.9 14.1   14.1   15.0
                          
  22, 23, 2 4:  'word' type annotations
  24 :  'event_01' type annotation


Note that back/circular arcs are not allowed in annotataion
graphs.  Annotation graph API doesn't prevent that though.

Can you think of other ways to do it?

Thanks,
Haejoong


On Fri, Oct 25, 2002 at 05:31:29AM -0700, Fredrik Karlsson wrote:
> Dear Dr. Lee,
> 
> I recently discovered the AG toolkit and found that it overlaps 
> substatially with something that I've just started to develop 
> for handling of my thesis data. Also, it provides a number of 
> features that I had opted against due to time limitations (such 
> as database connectivity).Therefore, using the already 
> developed AGTK would obviosly save me a lot of time.
> 
> However, I have two questions that I feel that I would like to 
> ask the development team before discontinuing my own 
> project.
> The first question addresses the need for labeling of events. 
> In acoustical phonetic analysis, there are quite a few 
> situations where the object in question is not a time interval, 
> but a point in time. Such example in phonetics is the onset of 
> voicing, turing points in the formant track or the articulatory 
> release of a plosive. How do you propose that this kind of 
> annotation should be handles in the AG library, given the 
> definition of the graph.
> 
> The second issue regards format extentions. In my project, I 
> use the Praat software since it's annotation format facilitates 
> the distinction between intervalls and points.
> Therefore, in order for me to use the AGTK, I would have to 
> have an object that parses the Praat files and inherits from 
> agfio. The thing is that I have already constructed a similar 
> object in Python (using regexp), which would mean that 
> writing such a module should be close to trivial.
> However, since I guess that I am not the only one interrested 
> in such an extension, it would probably be better if that 
> module would be developed within the development team of 
> AGTK. The quality and consistency would probably increase, 
> and changes in the API would probably be easier to transfer 
> to all modules that way.
> Therefore, if you think that it is a good idea to include 
> support of the Praat program to AGTK, I suggest that I send 
> you Perl/Python code (from my project) including regexp in 
> order for you to be able to include support for Praat 
> annotations in AGTK. That way, I and many others would be 
> able to use the AG library for the Praat format as well.
> 
> Please tell me what the development team think about this.
> 
> Yours truely,
> 
> /Fredrik Karlsson