- From: Felix Sasaki <fsasaki@w3.org>
- Date: Fri, 31 Aug 2012 10:40:19 +0200
- To: public-multilingualweb-lt@w3.org
- Message-ID: <CAL58czpxBv59j4rOb_cje7dS7vwCVM=vb_p_oQunLM54AJzsTQ@mail.gmail.com>
are at http://www.w3.org/2012/08/30-mlw-lt-minutes.html and below as text.
- Felix
[1]W3C
[1] http://www.w3.org/
- DRAFT -
MLW-LT WG
30 Aug 2012
[2]Agenda
[2] http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Aug/0319.html
See also: [3]IRC log
[3] http://www.w3.org/2012/08/30-mlw-lt-irc
Attendees
Present
arle, felix, yves, ankit, jirka, des, guiseppe, michael,
tadej, pablo, leroy, dave, milan, phil, david
Regrets
Jörg
Chair
felix
Scribe
daveL
Contents
* [4]Topics
1. [5]ACTIONS
2. [6]PRAGUE + TPAC
3. [7]ISSUE-22
4. [8]ISSUE-34 quality data category
5. [9]readibility of the spec
6. [10]ISSUE-36 special requirements
7. [11]ISSUE-41 mtConfidence
8. [12]time of this call
* [13]Summary of Action Items
__________________________________________________________
ACTIONS
<fsasaki> action-164?
<trackbot> ACTION-164 -- Maxime Lefrançois to create an RDF
ontology for ITS, and make sure that we update the ITS XSD to
accomodate the RDF output created via ACTION-147 -- due
2012-08-16 -- OPEN
<trackbot>
[14]http://www.w3.org/International/multilingualweb/lt/track/ac
tions/164
[14] http://www.w3.org/International/multilingualweb/lt/track/actions/164
tadej; followed this but can't dedicate time
<scribe> ACTION: felix to liaise with maximie to complete this
when time permits [recorded in
[15]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action01]
<trackbot> Created ACTION-205 - Liaise with maximie to complete
this when time permits [on Felix Sasaki - due 2012-09-06].
<fsasaki> action-179?
<trackbot> ACTION-179 -- David Lewis to work with Olaf-Michael
to help draft blurb for international organizations. -- due
2012-08-16 -- OPEN
<trackbot>
[16]http://www.w3.org/International/multilingualweb/lt/track/ac
tions/179
[16] http://www.w3.org/International/multilingualweb/lt/track/actions/179
<fsasaki> dave: talked about this with olaf, need to follow up
<fsasaki> .. ongoing
<fsasaki> action-196?
<trackbot> ACTION-196 -- Felix Sasaki to make inheritance -
overriding etc. clearer in the spec -- due 2012-08-20 -- CLOSED
<trackbot>
[17]http://www.w3.org/International/multilingualweb/lt/track/ac
tions/196
[17] http://www.w3.org/International/multilingualweb/lt/track/actions/196
<fsasaki>
[18]http://lists.w3.org/Archives/Public/public-multilingualweb-
lt/2012Aug/0311.html
[18] http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Aug/0311.html
<fsasaki>
[19]http://www.w3.org/International/multilingualweb/lt/drafts/i
ts20/its20.html#datacategories-defaults-etc
[19] http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#datacategories-defaults-etc
felix: inheritence and overriding colum are mutually redundant
and make the function of inhertance clearer
... does everyone agree?
Yves; in broad agreement
<fsasaki> dave: old text didn't really say what overriding did
<fsasaki> .. there is several aspects like "overriding the data
values of the ITS category"
<fsasaki> .. that might make it clearer
<fsasaki> .. so there wasn't a definition what is overriden
<fsasaki> .. the data values that apply to the nodes get
overriden
PRAGUE + TPAC
<scribe> scribe: daveL
felix: hope eveyone can come to both, though prague is most
important
... but TPAC also give opportunity to meet other W3C folk
Des; will be showing some implementations
felix; people are implementing and we need slot in agenda for
this
<Yves_> i should have some data categories working and useable
too (XLIFF 1.2 and 2.0 output for example)
daveL: will have some integration between CMS-LION and mose MT
and perhaps XLIGG
felix: ask eveyone to add any plans for demonstrations to
prague agenda page
<fsasaki>
[20]http://www.w3.org/International/multilingualweb/lt/wiki/Pra
gueSep2012#Objectives
[20] http://www.w3.org/International/multilingualweb/lt/wiki/PragueSep2012#Objectives
<fsasaki> "Your implementations and use cases here: ..."
<scribe> ACTION: felix to follow up on demo commitments at
prague under objectives on pargue agenda page [recorded in
[21]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action02]
<trackbot> Created ACTION-206 - Follow up on demo commitments
at prague under objectives on pargue agenda page [on Felix
Sasaki - due 2012-09-06].
felix: you can add a point to page showing more details also.
<fsasaki> dave: people can be free to set up a page for "their"
implementation to put more details in
<fsasaki> felix: sure
ISSUE-22
felix; ISSUE 22 provenance, need update
<fsasaki> dave: working on it. For actual text need to add
pointers to translation agent
<fsasaki> .. where was the issue related to mtScore
<fsasaki> .. standoff just needed the pointer version added
<fsasaki> .. felix had sent a comment about "standoff prov.
MUST point to a provenance record"
<fsasaki> .. how would you test conformance on that
<fsasaki> .. is it sufficient to have a URL and get a URL back
<fsasaki> felix: how would validation work?
<fsasaki> dave: several options for that in provenance
<fsasaki> .. basically a GET operation and you get something
back in a certain format
<fsasaki> .. if you have a pointer to one of the records
<fsasaki> .. they need to follow a link and do an HTTP get
<fsasaki> .. to retrieve it
<fsasaki> .. but do they have to parse the record
felix: you can have ITS data category conformance by access the
URL, but there is more to showcasing functionality
... so test suite is artificial, but we want to have more live
tests
... so we want the functional testing to really show the
usefulness
... test can cover both, we can do both at the same time
... so current test suite isn't useful for real-life
implementations
<Ankit> will having a MT system translate ITS metadata (as in
testsuite) tagged text be an example of functional testing?
dave: suggest implementers provide thier funcitonal test when
demoing in prague
yves; can provide some examples in prague
flix; to get some feedback on test suite input and output from
HTML WG
<scribe> ACTION: felix to get some feedback on test suite input
and output from HTML WG [recorded in
[22]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action03]
<trackbot> Created ACTION-207 - Get some feedback on test suite
input and output from HTML WG [on Felix Sasaki - due
2012-09-06].
ISSUE-34 quality data category
<fsasaki>
[23]http://www.w3.org/International/multilingualweb/lt/drafts/i
ts20/its20.html#lqissue
[23] http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#lqissue
<fsasaki>
[24]http://www.w3.org/International/multilingualweb/lt/drafts/i
ts20/its20.html#lqprecis
[24] http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#lqprecis
felix: yves has update specification as above
Des: did quality score issue get resolved
... there were 4 options mentioned
<fsasaki> see mail here
[25]http://lists.w3.org/Archives/Public/public-multilingualweb-
lt/2012Aug/0267.html
[25] http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Aug/0267.html
yves: issue of whether it is a range or a vote need to be
resolved, and an attribute added if both to be supported
Des: can live without voting if it overcomplicate things
Yves: lets check if other partners such as microsoft have an
interest
Des: it is a valid use case
Yves: yes, but it should be an option in that data category
rather than a new data category
Des: agrees
Yves; volunteer to add a proposal so we can explore the issues
<scribe> ACTION: yves to add an option to quality precis draft
supporting voting ranges [recorded in
[26]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action04]
<trackbot> Created ACTION-208 - Add an option to quality precis
draft supporting voting ranges [on Yves Savourel - due
2012-09-06].
readibility of the spec
<fsasaki> .. currrently trying to re-draft parts of section 1
<fsasaki> .. disambiguation is very hard to read
<fsasaki> .. using a lot of computational linguistic
terminology
<fsasaki> .. propose not to change a lot, but the prose
<fsasaki> tadej: posted a new version a while ago, aware of
this problem
<fsasaki> .. whole data category talks about computational
linguistic domain
<dF> perhaps tadej should mute
felix: want to have some people to review some data category
text with the eye of someone form a different domain
<scribe> ACTION: tadej to review readability and make proposals
to improve readability to non-localisation people [recorded in
[27]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action05]
<trackbot> Created ACTION-209 - Review readability and make
proposals to improve readability to non-localisation people [on
Tadej Štajner - due 2012-09-06].
<scribe> ACTION: daveL to review diambiguation from the point
of view of readability [recorded in
[28]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action06]
<trackbot> Created ACTION-210 - Review diambiguation from the
point of view of readability [on David Lewis - due 2012-09-06].
<fsasaki>
[29]http://www.w3.org/International/multilingualweb/lt/drafts/i
ts20/its20.html#lqprecis
[29] http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#lqprecis
<scribe> ACTION: Michael to review quality precis for
readability from non-localisation viewpoint [recorded in
[30]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action07]
<trackbot> Created ACTION-211 - Review quality precis for
readability from non-localisation viewpoint [on Michael Kruppa
- due 2012-09-06].
felix; also ask Dave, David and moritz to do a general review
of the draft
<fsasaki> dave: some data catgories have examples that do
something which is not explained. You need to explain: here is
an example how an element is selected, given the values etc.
<fsasaki> .. some of Yves ones are quite good
<fsasaki> .. I'd encourage everybody to look into that
ISSUE-36 special requirements
felix: summarised changes to special requirements and yves to
include this
action; yves to add allowed characters and stroage size to
specification
<dF> Happy with "forbidden characters" and "storage size"
solutions as per e-mail discussion..
<scribe> ACTION: Yves_ to add allowed characters and stroage
size to specification [recorded in
[31]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action08]
<trackbot> Sorry, couldn't find user - Yves_
ISSUE-41 mtConfidence
dF: harmonisation issue to be resolved, e.g. with provenance
... need more granularity on translation agent for this to
advance
<scribe> ACTION: daveL to call David to advance the transagent
and mtconfirdence score harmonisation [recorded in
[32]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action09]
<trackbot> Created ACTION-212 - Call David to advance the
transagent and mtconfirdence score harmonisation [on David
Lewis - due 2012-09-06].
dF: ask tadej what score format is relevant to text analytics
tadej: either 0 to 1 or percentage would work, but 0-1 would
work and is perhaps more common
<Michael> sorry, gotta run...cheers
tadej: summarize is 0 to 1 best solution
... this is simpler to process and percentage and can be XSD
float
dF: agree we will go for this option
time of this call
<fsasaki> would it work for everybody to have it one our
earlier?
<fsasaki> phil: would be OK
<Ankit> works for us too (DCU)
dF: could yves move earlier
yves: yes, can try
<fsasaki> ACTION: felix to send info about time change proposal
to the list [recorded in
[33]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action10]
<trackbot> Created ACTION-213 - Send info about time change
proposal to the list [on Felix Sasaki - due 2012-09-06].
<fsasaki> time propsosal - one hour later
<Ankit> Bye..
<gderiard> Good bye....
Summary of Action Items
[NEW] ACTION: daveL to call David to advance the transagent and
mtconfirdence score harmonisation [recorded in
[34]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action09]
[NEW] ACTION: daveL to review diambiguation from the point of
view of readability [recorded in
[35]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action06]
[NEW] ACTION: felix to follow up on demo commitments at prague
under objectives on pargue agenda page [recorded in
[36]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action02]
[NEW] ACTION: felix to get some feedback on test suite input
and output from HTML WG [recorded in
[37]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action03]
[NEW] ACTION: felix to liaise with maximie to complete this
when time permits [recorded in
[38]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action01]
[NEW] ACTION: felix to send info about time change proposal to
the list [recorded in
[39]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action10]
[NEW] ACTION: Michael to review quality precis for readability
from non-localisation viewpoint [recorded in
[40]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action07]
[NEW] ACTION: tadej to review readability and make proposals to
improve readability to non-localisation people [recorded in
[41]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action05]
[NEW] ACTION: yves to add an option to quality precis draft
supporting voting ranges [recorded in
[42]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action04]
[NEW] ACTION: Yves_ to add allowed characters and stroage size
to specification [recorded in
[43]http://www.w3.org/2012/08/30-mlw-lt-minutes.html#action08]
[End of minutes]
__________________________________________________________
Minutes formatted by David Booth's [44]scribe.perl version
1.136 ([45]CVS log)
$Date: 2012/08/31 08:34:16 $
[44] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
[45] http://dev.w3.org/cvsweb/2002/scribe/
Received on Friday, 31 August 2012 08:40:49 UTC