1 Agenda of LT-Web Kick-off Meeting 18-19 January 2012

Participants: See separate list of participants.

IRC channel: #multilingualweb-lt

IRC Web client: go to http://irc.w3.org, enter your nickname and the channel #multilingualweb-lt

Presentations

1.1 Afternoon 18th January (starting 1 p.m.) At DFKI

Location: http://www.dfki.de/web/contact/dfki-project-office-berlin?set_language=en&cl=en

Presentations about projects in the context of LT-Web, demos etc. Topics / demos / discussions:

• 13:00 Welcome (Hans Uszkoreit + Felix Sasaki)
• 13:15 Demo / presentations slot 1
  • 13:15 – 13:35 MaTrEx system: premises and promises for using metadata (Declan Groves)
  • 13:35 – 13:55 Passing metadata from DocBook to HTML (Jirka Kosek)
  • 13:55 – 14:15 Localization metadata in the Okapi framework (Yves Savourel)
  • 14:15 – 14:35 Producing metadata via Enricher (Tadej Štajner)
• 14:45 Break
• 15:15 Demo / presentations slot 2
  • 15:15 – 15:35 Linguaserve localization tool chain (Pedro L. Díez Orzas)
  • 15:35 – 15:55 SOLAS system and metadata in XLIFF (David Filip)
  • 16:15 – 16:35 META-NET / META-SHARE introduction (Georg Rehm)
  • 16:35 – 16:55 Modelling across content management, localization and MT processes (Dave Lewis)
  • 16:55 – 17:30 (Buffer slot, or early wrap up) General discussion, including members from the DFKI LT lab

1.2 Social event including dinner starting after the wrap up

• After the Wednesday meeting we will go the museum of computer games, see

http://www.computerspielemuseum.de/index.php?lg=en&sid=b4ee6bc402bd6a47baff48a0a976eb6d and http://tinyurl.com/computerspielemuseum

• We will have a one hour guided tour.
• After that tour we will go to the Berlin TV Tower http://www.tv-turm.de/en/index.php and have a three course menu, watching Berlin from the top.

NOTE: We will get around the city (DFKI offices > museum > TV Tower) with public transportation. I will have a group ticket, so just stay with me.

• After the dinner you can join me to go back to the DFKI office, or take a taxi to your hotel.

1.3 19th January

Location: Hotel Adrema http://www.hotel-adrema.de/default.aspx?lang=en http://www.hotel-adrema.de/map.aspx?lang=en

Topics:

• 9:00 Welcome and Technicalities
• 9:15 Report of the findings from 18th January (Felix Sasaki)
• 9:30 How LT-Web fits into the big picture (Kimmo Rossi)
• 10:00 LT-Web and the META-NET strategic research agenda (Georg Rehm)
• 10:30 Break & Group Photo (Richard Ishida)
• 11:15 Continuation of the MLW brand (Richard Ishida, Felix Sasaki)
  • Name of the Working Group: MLW-LT-LT-Web- ...
  • Name of the technologies to be developed: ITS 2.0, XYZ 1.0, ...
  • MultilingualWeb-LT communication strategy (Arle Lommel): key speaking points, who to reach out too
• 12:30 Lunch break
• 13:30 Upcoming events
• 14:45 W3C Infrastructure and LT-Web internal infrastructure: project internal communication, document store, etc. (Felix Sasaki)
• 15:00 Further demos / presentations
  • Multilingual CMS (Moritz Hellwig)
  • Moses MT System (David Vilar Torres)
  • M4loc (Milan Karasek and Yves Savourel)
  • VistaTEC (Phil Ritchie)
• 17:00 Administrative & Financial Matters, Portfolio of other projects from the current call (Kimmo Rossi)
• 17:30 Wrap up

1.4 Minutes / findings

Raw minutes of 18 January

Findings:

• The project can play an important role in "selling" language technology to different audiences: users, industries, etc. (Hans)

• On the technical level, various pieces can be integrated via our metadata:

• • localization tools (Yves) or MT tools (Declan) make use of automatically generated metadata (tadej), e.g. for term identification / glossary generation
  • From the deep web metadata can be passed to the surface web via standardized transformations - example DocBook > HTML (Jirka)
  • localization tool chains can be integrated with CMS (Pedro, related to Moritz)

• Various types of metadata is needed, see the charter draft and

• • revision information (Pedro)
  • domain (or genre?) information (Declan and Georg)
  • Distributed service integration (David)

• Language resources need metadata to be found and to be distributed (Georg)

• Though above aspects are important for our technical work, the success of the work depends on a branding strategy:

• • How to make more money with language related technologies (Georg, Kimmo, Dag, ...)

• • Make sure that the message is not only specific to Europe (Richard)

• • Reach out to end users of the technology, i.e. people working with metadata, SMEs relying on it using re-usable systems, reduce amount of human intervention in translation process (Pedro)

• • Integrate with software infrastructure (Dag)

• • Involve also other communities working in (natural language and other) types of data, e.g. Semantic Web (Dave)

• Important next steps are:

• • Decide about the branding to the real world, including a general MultilingualWeb branding

• • Decide about the branding means (installation of the W3C working group including press releases, Gala Monaco presentation) and move them forward

• • Develop requirements bottom up: what do existing implementations do, how can we bring them closer together and document that? Example Enrycher vs. Okapi framework

• • Involve the necessary communities: 1st workshop, including MultilingualWeb in general