SW-HamNoSys

[Sarah Ebling]

Dear Gerard,

thanks a lot for your message. 

We are converting from SW to HNS only, not the other direction (for the reason you mentioned). 

"Frequency" refers to the frequency of co-occurrence of individual SW and HNS symbols. Our approach works as follows: first we manually map SW categories to HNS parameters, e.g., the SW "Hands" category (http://www.signbank.org/iswa/) to the handshape parameter in HNS. We do this only once. Then, for each sign for which we have both an SW and a HNS notation we automatically pair each individual SW and HNS symbol, obeying the restrictions we established, i.e., we only pair an SW symbol with a HNS symbol if they belong to a category/parameter that we mapped beforehand. This will give us a list of individual HNS symbol correspondences for each individual SW symbol, ranked by frequency. The idea is that the most frequent individual HNS symbol corresponding to an individual SW symbol is a good candidate for an actual HNS correspondence.  Of course, this approach gets better as we get more data. We currently have about 1200 signs for which both SW and HNS notations exist.

We'd be happy to receive additional feedback from you -- or from anyone else out there.

Best wishes,
Sarah and Penny

>> Anfang der weitergeleiteten E-Mail:
>> 
>>> Von: "Gerard Meijssen" <gerard.meijssen at GMAIL.COM>
>>> Betreff: Re: SW-HamNoSys
>>> Datum: 29. Juni 2012 08:56:33 MESZ
>>> An: SLLING-L at LISTSERV.VALENCIACOLLEGE.EDU
>>> Antwort an: "linguists interested in signed languages" <SLLING-L at LISTSERV.VALENCIACOLLEGE.EDU>
>>> 
>>> Hoi,
>>> What I do not understand is how you can do a frequency analysis and make it work. I expect that you concentrate on one sign language. SignWriting is used by many sign languages and consequently the logic escapes me. It is like comparing the Latin script and the Arabic script; you cannot do a two way transcription it is impossible because of the architecture of the two scripts. Likewise with SW and HS; HS is the poor cousin who lacks many features SignWriting provides.
>>> Thanks,
>>>       Gerard
>>> 
>>> On 28 June 2012 16:55, Penny Boyes Braem <boyesbraem at fzgresearch.org> wrote:
>>> Hallo all,
>>> 
>>> We just wanted to thank everyone who has contributed comments on this list to our previous email concerning our pilot project for SignWriting-HamNoSys notation conversion.  As a result of the posting, we’ve had helpful correspondence with some groups who are also interested in working on conversion between different types of sign notation or coding systems.   (Just to be clear, we are working with individual lexical items and not a corpus of signed texts.)
>>> 
>>> Clearly, the fact that there are more SW than HNS symbols will make any conversion not straightforward.  The non manuals are also a special concern.  At the moment, we’re hoping that some of the new two letter codes for HNS manuals developed by the recent EU Visicast project will be of help here.   
>>> 
>>> In our pilot project, we are  trying to establish mappings between SW and HNS symbols (or combination of symbols)  based on frequencies of occurrences - i.e. a ‘bottom-up’ approach.  A  ‘first pass’ comparison we are currently making will help us focus on those areas of notation which pose particular problem for conversions, and on which we will concentrate in the next steps.
>>>  
>>> We’d be happy to share our first pass results with any of you who are interested - please contact us at our individual emails.
>>> 
>>> Thanks again to all,
>>> Penny & Sarah 
>>> 
>>> Penny Boyes Braem (boyesbraem at fzgresearch.org) and  Sarah Ebling (ebling at cl.uzh.ch)
>>> 
>>> Dr. Penny Boyes Braem
>>> boyesbraem at fzgresearch.org
>>> www.fzgresearch.org
>>> Switzerland
>>> 
>>> 
>>> Dr. Penny Boyes Braem, Director
>>> Forschungszentrum für Gebärdensprache (FZG)
>>> (Center for Sign Language Research)
>>> CH-4059 Basel, Switzerland
>>> 
>>> boyesbraem at fzgresearch.org
>>> www.fzgresearch.org
>>> 
>>> 
>>> 
>>> 
>> 
>> Dr. Penny Boyes Braem, Director
>> Forschungszentrum für Gebärdensprache (FZG)
>> (Center for Sign Language Research)
>> CH-4059 Basel, Switzerland
>> 
>> boyesbraem at fzgresearch.org
>> www.fzgresearch.org
>> 
>> 
>> 
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/slling-l/attachments/20120629/82d41ada/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2546 bytes
Desc: not available
URL: <http://listserv.linguistlist.org/pipermail/slling-l/attachments/20120629/82d41ada/attachment.bin>