Re: varstrans.owl

Dear all,

  I had a f2f discussion with John this morning and I must modify my 
proposal from yesterday. I agree that the translation relation in 
dictionaries is by definition a relation between senses, as in any 
standard dcitionary you have translations indicated for every sense of 
the lexical entry in question. So I agree then that we should keep the 
term "translation" as a relation between senses.

*However*, John and myself see the use case for having a relation that 
states that a lexical entry is translatable into some other lexical 
entry in *some* context and under *some* interpretation (sense) of the 
lexical entry, so our proposal is to introduce a relation 
"translatableAs" defined between LexicalEntries.

I think that nicely addresses the use case and sovles our problems.

So unless anybody objects, I would proceed to have those two relations:

1) translation: defined between senses of lexical entries in different 
relations, corrresponding to the standard relation in bilingual dictionaries
2) translatableAs: defined between lexical entries that *can* be the 
result of translating one into the other (in some not further 
specificied context or sense).

Happy holidays.

Philipp.

Am 31.07.14 22:48, schrieb John P. McCrae:
> Hi Philipp,
>
> As you may have guessed I don't agree with this. We are not working 
> with strings we are defining a lexical model and as such the first two 
> definitions are not applicable to our domain, that is we don't work 
> with strings we work with senses and entries and forms! By your logic, 
> translation is actually a relationship between forms not entries, as 
> the form "Katze" translates to "cat" and the form "Katzen" translates 
> to "cats".
>
> The idea of calling the 'translation' property 
> 'crosslingualequivalentsense' would be in opposition to standard 
> lexicographic practice as well as inconsistent with most existing 
> lexical resources. Furthermore, it encourages the confounding practice 
> of putting translation between lexical entries.
>
> It is clear (at least to me) that translation is a property that must 
> involve the meaning of the word, and not including the sense would 
> undermine the robustness and usability of resources created with the 
> model by condoning ambiguous cross-lingual linking between resources.
>
> Regards,
> John
>
>
> On Thu, Jul 31, 2014 at 10:11 PM, Philipp Cimiano 
> <cimiano@cit-ec.uni-bielefeld.de 
> <mailto:cimiano@cit-ec.uni-bielefeld.de>> wrote:
>
>     Dear all,
>
>      and the last email for today. I would like to propose a
>     terminological change for the vartrans module.
>
>     We had a lot of discussion about the notion of translation a few
>     months ago, showing that the notion of translation is indeed quite
>     ambiguous, between:
>
>     1) the process of translation a source string into a target string
>     2) the result of this process
>     3) the relation between two "equivalent" senses in to different
>     languages
>
>     I would like to propose a deeper change to the vartrans module,
>     that is the one of using "translation" only for the result of
>     translating some source string into a target string as an opaque
>     relationin some context, under some conditions etc.
>
>     If we want to be technically precise and say that there is a
>     relation of cross-lingual translational equivalence between two
>     senses, then we should use a more technical relation such as
>     "CrossLingualEquivalentSense".
>
>     Many people use "translation" as an opaque relation denoting the
>     result of translating one string into another and are puzzled by
>     the fact.
>
>     If we want to be technically precise and say that translational
>     equivalence is a relation between senses and not lexical entries,
>     then we should also use a more technical term such as
>     "CrossLingualEquivalentSense" and leave the more vague
>     "translation" term for the relation between two lexical entries
>     that can be translated into reach other in *some* context.
>
>     Sorry for opening this discussion again, but I believe it is for
>     the robustness and usability of the model to rethink this at least
>     once more. It is now the right moment given that I am aiming to
>     finalize the vartrans module in the near time.
>
>     Opinions?
>
>     Philipp.
>
>     -- 
>     --
>     Prof. Dr. Philipp Cimiano
>     AG Semantic Computing
>     Exzellenzcluster für Cognitive Interaction Technology (CITEC)
>     Universität Bielefeld
>
>     Tel: +49 521 106 12249 <tel:%2B49%20521%20106%2012249>
>     Fax: +49 521 106 6560 <tel:%2B49%20521%20106%206560>
>     Mail: cimiano@cit-ec.uni-bielefeld.de
>     <mailto:cimiano@cit-ec.uni-bielefeld.de>
>
>     Office CITEC-2.307
>     Universitätsstr. 21-25
>     33615 Bielefeld, NRW
>     Germany
>
>
>

-- 
--
Prof. Dr. Philipp Cimiano
AG Semantic Computing
Exzellenzcluster für Cognitive Interaction Technology (CITEC)
Universität Bielefeld

Tel: +49 521 106 12249
Fax: +49 521 106 6560
Mail: cimiano@cit-ec.uni-bielefeld.de

Office CITEC-2.307
Universitätsstr. 21-25
33615 Bielefeld, NRW
Germany