SignWriting to HamNoSys?

[Penny Boyes Braem]

Thank you all (especially Maria and Franz) for your thoughts these past couple of weeks on topics concerning  SW-HNS conversion.  Again, we’d like to emphasize that our project is a small beginning step in this direction, for which we are first concentrating on converting Swiss German Sign Language (DSGS) citation forms of signs (not signs from a text), in order to see if there are any patterns which would encourage us to further pursue this attempt. As a second step in this small exploratory study, it would be helpful to compare our findings for HNS with a set of  signs from another sign language, which are also already transcribed with both SW and HNS.  
Franz, your comments on the different requirements and aims of notations  is a helpful reminder that if we are able to compare the DSGS results with those from other sign languages, we absolutely need to notate the aims (and levels) of all sets of notations in our metadata.
Thanks again, and we’ll keep you informed of our progress.
Penny and Sarah



Am 24.08.2012 um 23:05 schrieb MARIA GALEA:

> Hello again all and especially Profs Dotter and Machado,
> 
> Thank you for your feedback. I really appreciate your comments. I would
> really like to read up more about your works. Do you think you could
> direct me to a few of your recent papers, especially anything that has to
> do with SignWriting.
> 
> Yes Machado - I remember meeting you in Malta at the LREC 2010 - if I rem.
> well you are from Brazil, am I right? I tried to search for your paper on
> the LREC 2010 website  but was unsuccessful. What is your full name if I
> may ask?
> 
> If the people undertaking the work of the SignWriting to HamNoSys
> converter are reading this, do you have any more details about the work
> that I can read about..i need more information if possible.
> 
> Thanks all and have a great weekend,
> maria
> 
> 
>> Hi, Maria!!!
>> 
>> very interesting issue pointed out by you.
>> Although Franz's answer was very sound in some aspects, I am not sure he
>> has addressed all of your points.
>> 
>> Anyway, what you both brings to surface, IMHO, is that one has to be
>> really
>> aware of the issue when dealing with such conversions: general
>> translation/conversion vs. language specific rules.
>> 
>> I would dare to say this could be a research topic in its own. But I would
>> like to bring the corpora based inference approach to the problem.
>> 
>> I do know we do not have many sign language corpora available in the two
>> notations (let alone annotated corpora or paralell SW | HNS | Gloss ones).
>> But I have had myself an initial experience in providing sign language
>> users ways to continuously grow their own corpus using some interesting
>> inference techniques. In fact, I even had the pleasure to meet Maria in
>> person, when I visited Malta to present the work in LREC 2010.
>> 
>> Anyway, my point is maybe it is really tough to deal with the generic vs.
>> specific issue using hardcoded rules and probably we shold look for ways
>> (fundings?) to develop sign language corpora built for both real life
>> usage
>> and research purposes.
>> 
>> Best
>> 
>> Machado
>> 
>> 2012/8/23 MARIA GALEA <maria.azzopardi at um.edu.mt>
>> 
>>> Hi everyone - first a v brief introduction - I have been using
>>> SignWriting
>>> for data transcriptions for several years and have also used it as a
>>> writing system to write long texts in Maltese Sign Language (for sign
>>> language literature - not for research).
>>> 
>>> I'm working on my dissertation 'SignWriting: Towards an orthography for
>>> LSM (Maltese Sign language)' with the Univ of Malta. I can't believe
>>> this
>>> email list (SSLING) never came to attention before. Better (extremely)
>>> late than never I guess.
>>> 
>>> Can I ask a few questions about the work on the SignWriting to HamNoSys
>>> converter: what will be the use of this tool? I.e. WHY are you carrying
>>> out this study? Why do you need an SW converter to HamNoSys? Are you
>>> focussing on one sign language or several?
>>> 
>>> Gerard's argument is valid - but besides this, there are different rules
>>> applied to SignWriting for different sign languages - although this is
>>> still not researched (I hope to be filling this gap a little), with time
>>> users of the SW writing system naturally adapt SignWriting.(In the same
>>> way that the latin script "x" symbol nowadays represents a wide varied
>>> list of sounds that is language dependant - and has moved away from the
>>> original Latin sound). There is evidence that with the use of long-text
>>> writing of SignWriting for a specific sign language, the system
>>> naturally
>>> evolves beyond a simple phonetic transciption (i.e. a representation of
>>> exact articulation) to a more abstract representation of language.  How
>>> do
>>> you hope to deal with this reality, when this reality has not yet been
>>> fully understood and researched?
>>> 
>>> Taking the analogy of say written English to I.P.A (International
>>> Phonetic
>>> Alphabet)- I could imagine an IPA converter dealing with the English
>>> word
>>> 'although', because I imagine you could 'teach' your machine the pattern
>>> of written English '-ough' : that represents the sound "ow". Most
>>> written
>>> English patterns are known so you could teach them to your machine I
>>> guess. But what about the SignWriting to HamNosys converter? How can you
>>> include such patterns, if they haven't yet been investigated? And
>>> furthermore how are you going to include all different patterns for all
>>> different sign languages that are using SignWriting?
>>> 
>>> Look forward to your reply - and wish you all a lovely day.
>>> Maria
>>> 
>>> 
>>> !DSPAM:5035d7ff210502100250580!
>>> 
>>> 
>>> 
>> 
> 

Dr. Penny Boyes Braem
Forschungszentrum für Gebärdensprache (FZG) 
and
Sarah Ebling
Center for Computer Linguistic, University of Zurch



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/slling-l/attachments/20120831/1d3feee4/attachment.htm>