W3C home > Mailing lists > Public > public-multilingualweb-lt-comments@w3.org > December 2012

Re: issue-60 (Re: Comment on ITS 2.0 specification WD)

From: Felix Sasaki <fsasaki@w3.org>
Date: Thu, 13 Dec 2012 16:13:06 +0100
Message-ID: <50C9F082.3060109@w3.org>
To: public-multilingualweb-lt-comments@w3.org
Hi Jörg, all,

I tried to implement this in the draft, see
http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#lqissue-typevalues

If there is no disagreement I would close the issue on the monday call.

Best,

Felix

Am 12.12.12 14:47, schrieb Jörg Schütz:
> Hi Arle,
>
> Some corrections and amendments for #1:
>
> (1) A text is defective in ways the defy categorization, ... => A text 
> is defective in ways to defy categorization, ...
>
> (2) (e.g., a translation shows severe grammatical defects and appears 
> unrelated to the source material) => (e.g., a translation shows an 
> unintelligible result and/or appears unrelated to the source material)
>
> Cheers -- Jörg
>
> On Dec 12, 2012 at 09:21 (UTC+1), Arle Lommel wrote:
>> If we take this approach, here is a pass at the information needed for
>> #1 with changes in *red bold*
>>
>> *Value*
>>
>>     uncategorized
>>
>> *Description*
>>
>>     The issue *either *has not been categorized *or cannot be 
>> categorized*
>>
>> *Example*
>>
>>   * A new version of a tool returns information on an issue that has not
>>     been previously checked and that is not yet classified.
>>   * *A text is defective in ways the defy categorization, such as the
>>     appearance of nonsense garbled text of unknown origin (e.g., a
>>     translation shows severe grammatical defects and appears unrelated
>>     to the source material)*
>>
>> *Scope*
>>
>>     S or T
>>
>> *Notes*
>>
>>     This category has two *the following* uses:
>>
>>       * A tool can use it to pass through quality data from another tool
>>         in cases where the issues from the other tool are not classified
>>         (for example, a localization quality assurance tool interfaces
>>         with a third-party grammar checker).
>>       * A tool's issues are not yet assigned to categories, and, until
>>         an updated assignment is made, they may be
>>         listed asuncategorized. In this case it is recommended that
>>         issues be assigned to appropriate categories as soon as possible
>>         since uncategorized does not foster interoperability.
>>       * *Uncategorized can be used where a portion of text is defective
>>         in a way that defies assignment to a category in either the
>>         originating system or in any other ITS localization quality
>>         markup to indicate that it is uncategorizable.*
>>
>> #2 would come along next year.
>>
>> #3 probably wouldn't need much update at this point since their is only
>> a slight expansion of meaning in this category. However, when QTLP's
>> tool develops I could add it in. This would again be next year.
>>
>> My guess, by the way, is that this can be seen as clarification of
>> usage, rather than a substantive change, but we can see what others 
>> think…
>>
>> -Arle
>>
>>
>>
>> On 2012 Dec 12, at 06:17 , Felix Sasaki <fsasaki@w3.org
>> <mailto:fsasaki@w3.org>> wrote:
>>
>>> Thank you, Jörg. Going the "stability path" seems also reasonable
>>> given this positive development
>>> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Dec/0061.html 
>>>
>>>
>>> So the actions needed would be
>>>
>>> 1) clarification of "uncategorized"
>>> 2) having an example that demonstrates the usage in the MT scenario -
>>> not necessarily in the spec, but as part of best practices and to see
>>> the annotation the qt launchpad project would produce
>>> 3) update
>>> http://www.w3.org/International/its/wiki/Tool_specific_mappings#Mappings_for_Localization_Quality_Issue_Type 
>>>
>>> http://www.w3.org/International/its/ig/its20-tool-specific-mappings.html 
>>>
>>>
>>> Arle, would that work for you? If yes, when could you do 1-3?
>>>
>>> With regards to Phil's mail at
>>> http://lists.w3.org/Archives/Public/public-multilingualweb-lt-comments/2012Dec/0010.html 
>>>
>>> I see this as a different topic, but would prefer not to add values or
>>> attributes at this time, like with issue-60. Phil, if you still need
>>> this please create a seperate comment.
>>>
>>> Best,
>>>
>>> Felix
>>>
>>> Am 11.12.12 20:57, schrieb Jörg Schütz:
>>>> That's a very good solution to avoid a possible type value tsunami
>>>> and additional LC (if this is really the case with such an addition).
>>>>
>>>> By the way, your "1862" example is a candidate for the
>>>> "mistranslation" type.
>>>>
>>>> Cheers -- Jörg
>>>>
>>>> On Dec 11, 2012 at 18:31 (UTC+1), Arle Lommel wrote:
>>>>> The other alternative is that we expand the semantics of 
>>>>> "uncategorized"
>>>>> slightly to allow for a more naturalistic interpretation such that it
>>>>> doesn't mean "we haven't categorized it" to "we haven't or can't
>>>>> categorize it". That would be satisfactory as well, I think, and 
>>>>> less of
>>>>> a change.
>>>>>
>>>>> -Arle
>>>>>
>>>>>
>>>>>
>>>>> On 2012 Dec 11, at 18:27 , Arle Lommel <arle.lommel@dfki.de
>>>>> <mailto:arle.lommel@dfki.de>> wrote:
>>>>>
>>>>>> Jörg is correct here that nothing has this already. This is really
>>>>>> looking forward to QT Launchpad work. If saying "nobody has
>>>>>> implemented this so far" disqualifies it, then we would be forced to
>>>>>> use "uncategorized" and add some custom markup. That wouldn't be the
>>>>>> end of the world for us, but it would be nice to have.
>>>>>>
>>>>>> However, see my last mail about how I see this as different.
>>>>>>
>>>>>> (I can say, up front, that if this isn't accepted I won't hold
>>>>>> anything up over it, so the moment this causes real problems, we can
>>>>>> drop it.)
>>>>>>
>>>>>> -Arle
>>>>>>
>>>>>> On 2012 Dec 11, at 18:15 , Jörg Schütz <joerg@bioloom.de
>>>>>> <mailto:joerg@bioloom.de>> wrote:
>>>>>>
>>>>>>> Hi Felix,
>>>>>>>
>>>>>>> Since an additional value doesn't actually harm the type list which
>>>>>>> certainly can be seen as open ended (but still backward 
>>>>>>> compatible),
>>>>>>> the need for a subsequent LC is questionable.
>>>>>>>
>>>>>>> Nevertheless, the proposed quality type value "unintelligible" for
>>>>>>> the described output case might be controversial because it does 
>>>>>>> not
>>>>>>> indicate/reflect a quality consideration as the other types in the
>>>>>>> list do. The QT Launchpad project should therefore consider to use
>>>>>>> "uncategorized" because this value might indicate the "trashy"
>>>>>>> quality.
>>>>>>>
>>>>>>> And TMK, I'm not aware of any language proofing technology that 
>>>>>>> uses
>>>>>>> "unintelligible" has a quality value.
>>>>>>>
>>>>>>> Cheers -- Jörg
>
Received on Thursday, 13 December 2012 15:13:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 13 December 2012 15:13:36 GMT