ITS and big data and Minutes ITS IG Call 2014-01-12 from Felix Sasaki on 2015-01-12 (public-i18n-its-ig@w3.org from January 2015)

From: Felix Sasaki <fsasaki@w3.org>
Date: Mon, 12 Jan 2015 17:48:38 +0100
To: public-i18n-its-ig <public-i18n-its-ig@w3.org>
Message-Id: <115D2872-4221-4D23-85D1-ABAB0374B343@w3.org>

Hi all,

ITS call minutes are here
http://www.w3.org/2015/01/12-i18nits-minutes.html <http://www.w3.org/2015/01/12-i18nits-minutes.html>
and below as text. A few items from the call

1) We discussed XLIFF and preserve space. People on the call agreed with Yves’ solution see the end of the discussion here
http://www.w3.org/2015/01/12-i18nits-irc#T16-29-08 <http://www.w3.org/2015/01/12-i18nits-irc#T16-29-08>
2) We discussed the upcoming FREME project (start date 1 February) and who it will contribute to new ITS (and ITS - XLIFF) implementations. The project will have a slot in the big data event next week in Brussels https://ec.europa.eu/digital-agenda/en/news/horizon-2020-ict-16-big-data-networking-day <https://ec.europa.eu/digital-agenda/en/news/horizon-2020-ict-16-big-data-networking-day>
3) We touched in the relation of ITS2 to big data. ITS2 „Text analysis“ can help to store information about entities - a need which is more an more important for dealing with (small or big) web content. See an application of „Text Analysis“ here
http://googleknowledge.github.io/qlabel/ <http://googleknowledge.github.io/qlabel/>
Again, related implementations will be developed in the FREME project.

Best,

Felix

[1]W3C

[1] http://www.w3.org/

- DRAFT -

ITS IG

12 Jan 2015

[2]Agenda

[2] http://lists.w3.org/Archives/Public/public-i18n-its-ig/2015Jan/0001.html

See also: [3]IRC log

[3] http://www.w3.org/2015/01/12-i18nits-irc

Attendees

Present
Ankit, Yves, fsasaki, philr, renatb, serge

Regrets
christian, david

Chair
felix

Scribe
fsasaki

Contents

* [4]Topics
1. [5]roll call
2. [6]action items
3. [7]big data networking day
4. [8]ITS and JSON
5. [9]XLIFF mapping
6. [10]other XLIFF TC feedback
7. [11]other topics
8. [12]next call
* [13]Summary of Action Items
__________________________________________________________

<scribe> scribe: fsasaki

roll call

checking attendance ....

[14]http://lists.w3.org/Archives/Public/public-i18n-its-ig/2015
Jan/0001.html

[14] http://lists.w3.org/Archives/Public/public-i18n-its-ig/2015Jan/0001.html

action items

[15]https://www.w3.org/International/its/ig/track/actions/open

[15] https://www.w3.org/International/its/ig/track/actions/open

close action-54

<trackbot> Closed action-54.

action-9?

<trackbot> action-9 -- David Lewis to Look at the XLIFF 2.0
change tracking module for provenance -- due 2014-05-30 -- OPEN

<trackbot>
[16]http://www.w3.org/International/its/ig/track/actions/9

[16] http://www.w3.org/International/its/ig/track/actions/9

felix: will ping Dave about action-9

action-50?

<trackbot> action-50 -- Serge Gladkoff to Draft outreach mail
for doc for others to re-use -- due 2014-07-23 -- OPEN

<trackbot>
[17]http://www.w3.org/International/its/ig/track/actions/50

[17] http://www.w3.org/International/its/ig/track/actions/50

close action-50

<trackbot> Closed action-50.

serge: did the announcement on linkedin
... on linkedin groups many postings, but very few unfold into
real discussions with many replies - linkedin changes the way
the groups are working
... people got the announcement, volume of reaction is unclear
... if we want to do wider email outreach I can do that
... but it is important to get a more concrete message in
... in ITS I saw other topics like business value and benefits
per data category
... I can do a broad outreach once this content is done
... it would be nice to ask s.t. - email that does not require
any reaction will not cause any
... so then "business value" content is done, we will do
announcement via linkedin etc.
... also gala standards list
... but also need to see some request, e.g. what do we ask?
feedback, request for further participation etc.
... we know that new calls are about big data, we need to
connect ITS to big data, and when it will be picked up more

big data networking day

[18]https://ec.europa.eu/digital-agenda/en/news/horizon-2020-ic
t-16-big-data-networking-day

[18] https://ec.europa.eu/digital-agenda/en/news/horizon-2020-ict-16-big-data-networking-day

felix: will present "FREME" project
... that has some relation to ITS as well

phil: we are partner in FREME project
... we are adding some additional facilities to ocelt
... that will allow us to consume FREME services output
... that includes terminology and entity annotation services
... as somebody is translating and reviewing, that will be
extra facilities that will add add. value to translators
services
... not all thing we add will be open source, but part of
FREME, core things will be available
... so it is about adding to the business proposition of
translation services - adding some things to translation that
would not be part of the standards workflow
... adding some things that will be ITS2 or XLIFF2 etc.
metadata

felix: FREME will fudn work on ITS2 in XLIFF module

yves: another topic taht could be related to ITS
... people are using more and more web services like JSON
output

ITS and JSON

yves: a lot of service output is carried by JSON
... had some discussion we coudl do in JSON to use ITS
... you easily can do something with JSONpath to mimic ITS
rules, that could be an area of interest as well

phil: could be interesting
... would like the web services endpoints to be restful

felix: would be nice to discuss things in W3C related to FREME,
will see what other FREME partners see

phil: I am part of LD4LT group, happy to do some connection in
that directoin too

XLIFF mapping

[19]https://lists.oasis-open.org/archives/xliff/201412/msg00045
.html

[19] https://lists.oasis-open.org/archives/xliff/201412/msg00045.html

" preserve Space "

yves: for the mapping of preserver space - for structural data
there is no issue
... for inline text that has to be preserved, there is an issue
... in XLIFF you can have overlapping elements
... that creates issues for ITS
... xml:space applies to the content of the element - so ITS
would apply to empty elements
... instead of that we would define specific attributes
... we looked into how to make that backward compatible
... but that is hard
... david drafted a list of fallback solutions
... they all had some issues, I proposed a diferent one
... summary is: if you have an inline code that should preserve
space, you do normalization during extraction
... anything that could be normalized should be normalized
... I had some text proposal and re-arrangement in the spec,
that is there the discussion is
... do we need an inline representation for preserve space?
... if you use the fallback there is no need for the attribute

felix: normalization during extraction is a way to avoid the
attribute?

yves: yes
... you create a preserve space around the complete output
... so there is no need to say anything inline

phil: sounds reasonable

felix: agreement on the call to have a way without inline
preserve space

discussion is here:
[20]http://www.w3.org/2015/01/12-i18nits-irc#T16-29-08

[20] http://www.w3.org/2015/01/12-i18nits-irc#T16-29-08

yves: during XLIFF TC discussion david was the only one not
agreeing

other XLIFF TC feedback

yves: some ITS data categories are overlapping with XLIFF
features
... like provenance
... that has not been addressed so far
... mapping is not completely done, mapping not either
... either we will have to limit the data categories or to push
the deadline

felix: are tehre are some data categoires that are very
important for you?

phil: everything we are using in ocelot is very well advanced
... if we do find anything we will find it very quickely
... language quality and provenacne are very well advanced

yves: in provenance and quality not everything is mapped yet,
there is some work here