berthold.crysmann at gmail.com
Fri Jan 26 16:30:05 CET 2018
my paper on reduplication (including infixation) just came out a few
months ago in morphology. As for infixation, you will see that it is a
matter of distance. If this is about root and pattern morphology, you
may consider some sort of autosegmental representation. Look at Bird and
Klein 1994 for a theoretical approach, and at my Hausa tone paper
(Journal of Language Modelling 2015) for how these things can be done
with LKB and friends. Are you using vocalised or unvocalised Arabic? If
you translate some of the input into feature structures, you may have an
easier time. If you operate with autosegmental structures you will
certainly reduce distance. As far as I remember root have between 3 and
4 consonants, so the distance from the edge is maximally 2. For
autosegmental approaches to work, you'd need the preprocessing machinery
offered by pet and ace.
Why don't you post some example of what exactly you have in mind.
On 19/01/18 12:07, Guy Emerson wrote:
> Dear Samia,
> Thanks for your message. As far as I'm aware, the morphology module
> in the LKB is relatively limited and does not directly give you a way
> to define infixes. However, I think it should be possible to get this
> kind of behaviour -- for example, Berthold Crysmann has used string
> unification to implement reduplication for a grammar of Hausa. I
> wouldn't be able to explain the details of this, but perhaps one of
> the other developers on this list will be able to give you an answer
> in the next few days.
> For the source code of the LKB, you can find information here:
> http://moin.delph-in.net/LkbTop <http://moin.delph-in.net/LkbTop> .
> The source code repository is here: http://svn.emmtee.net/lkb/ . For
> information about a fully open source version of the LKB (in the sense
> that it does not require any proprietary compilers, runtime
> environments or code libraries), see: http://moin.delph-in.net/LkbFos
> 2018-01-19 9:33 GMT+00:00 Samia Ben Ismail <samia_benismail at yahoo.fr
> <mailto:samia_benismail at yahoo.fr>>:
> I'm Samia Ben Ismail. I'm from Tunisia. I'm a phd student. I use
> LKB to generate an Arabic linguistic resources. Can you tell
> me please it is possible to add new operations ( replacement, add
> and delete a letter) to the infix (middle) of the word. If, it is
> possible, tell me please, how I can obtain the open source of LKB?
> Samia Ben Ismail
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the developers