CUSTOMIZING MACHINE TRANSLATION
Draft Agenda Workshop #3 25 April 2019
Hilton Brussels Grand Place, Carrefour de l’Europe, Brussels
10:00 – 10min Luc Meertens Welcome
10:10 – 25min Philippe Gelin Objectives of DG Connect
10:35 – 25min Andreas Eisele Update of status eTranslation
11:00 – 15min Coffee Break
FOD Kanselarij + Joachim van den Sharing public services data to obtain
11:15 – 25min
Bogaert better MT quality
11:40 – 25min CrossLang, Sara Szoc MT Training Workflow
Human and automatic evaluation of
12:05 – 25min LT3, Lieve Macken
machine translation output
12:30 – 60min Lunch Break
13:30 – 25min Tilde, Andrejs Vasiljevs Terminology within MT
13:55 – 25min ELDA, Khalid Choukri Publicly available corpora
14:20 – 25min CrossLang, Tom Vanallemeersch Tools for Data Gathering – DIY
14:45 – 15 min Coffee Break
Round Table Discussion with
15:00 – 60min participation of DSIs, moderated by How to proceed from here?
Luc Meertens
16:00 End of workshop
Philippe Gelin, Objectives of DG Connect
The overall objective is to contribute to the development of CEF Automated Translation as a “multilingualism
enabler” for CEF DSIs, online services linked to CEF DSIs and other relevant public online services.
The specific objectives are (i) to gather information on additional needs of the CEF DSIs and public services;
(ii) to analyze the range of services which could extend CEF Automated Translation and (iii) to support CEF
DSIs and related systems with a view to maximizing their use of CEF Automated Translation services.
Andreas Eisele, Update of status eTranslation
FOD Kanselarij + Joachim van den Bogaert, Sharing public services
data to obtain better MT quality
In its-day-to-day activities, the translation department of the Belgian Chancellery of the Prime Minister
covers a wide range of topics. For some of these topics, specialized MT systems would be helpful (for
example, documents related to legal or policy matters), while for other topics, a broad-domain engine would
be more suitable (for example, press releases). In this talk, we explore the differences between broad-
domain and domain-specific translation from both the user and provider perspective. We discuss how public
services can benefit from tailored MT solutions, and how they can contribute to the development of better
MT systems at a European level.
Sara Szoc, MT Training Workflow
This presentation deals with the typical workflow for training an MT system, i.e incremental training. The MT
system is regularly being retrained as more and more parallel data become available (such as a growing
translation memory). Previous versions of the MT system act as a baseline which is improved upon, in terms
of translation quality. The concept of baseline also applies in case of domain-specific MT: the latter should
improve upon the general-purpose MT system.
Lieve Macken, Human and automatic evaluation of machine translation
output
This talk compares human and automatic evaluation methods. Automatic evaluation metrics are typically
used during the development of machine translation systems, for example to quickly compare successive
versions of a single system with each other. Human evaluation of MT output is highly informative, but it is
expensive in terms of time and expert human effort and may suffer from a lack of consistency.
Andrejs Vasiljevs, Terminology within MT
Khalid Choukri, Publicly available corpora
Tom Vanallemeersch, Tools for Data Gathering - DIY
While parallel corpora are available to a limited extent, much more parallel information can be found in
document archives or online resources, for instance inside multilingual web sites. This presentation discusses
tools and procedures that allow for automatically detecting equivalent documents or web pages and for
linking equivalent sentences within these pairs of documents or pages. Such equivalent sentences allow, for
instance, for the creation of a domain-specific parallel corpus.
Luc Meertens, Round Table Discussion “How to proceed from here ?”
Showcasing Language Technology in H2020 and
CEF Telecom projects
Draft Agenda Workshop #4 26 April 2019
Hilton Brussels Grand Place, Carrefour de l’Europe, Brussels
09:55 – 5min Luc Meertens, Consortium Welcome
10:00 – 15min Philippe Gelin / Alexandru Ceausu Objectives of DG Connect
10:15 – 15min Josef Von Genabith (DFKI) European Language Grid (H2020)
10:30 – 15min Andrejs Vasiljevs (Tilde) EU Council Presidency (CEF Telecom)
10:45 – 15min Zaineb Chelly (Inria) COMPRISE (H2020)
11:00 – 15min Coffee Break
Matthew Purver (Queen Mary University
11:15 – 15min EMBEDDIA (H2020)
of London)
11:30 – 15min Andrejs Vasiljevs (Tilde) eTranslation Termbank (CEF Telecom)
11:45 – 15min Manuel Herranz (Pangeanic) NEC TM Data (CEF Telecom)
12:00 – 15min Sara Szoc (CrossLang) APE-QUEST (CEF Telecom)
Paul Buitelaar (National University of
12:15 – 15min Prêt-a-LLOD (H2020)
Ireland Galway)
ParaCrawl – Corpus Gathering (CEF
12:30 – 15min Speaker of consortium tbd
Telecom)
12:45 – 45min Lunch Break
13:30 – 15min Thierry Etchegoyhen, Vicomtech ELRI (CEF Telecom)
13:45 – 15min Kenneth Heafield (Univ. of Edinburgh) BERGAMOT (H2020)
14:00 – 15min Miguel Angel Gómez Zotano (Everis) IADAATPA (CEF Telecom)
14:15 – 15min Speaker of consortium tbd ELITR (H2020)
Tamas Varadi (Hungarian Institute for
14:30 – 15min MARCELL (CEF Telecom)
Linguistics)
14:45 – 15min Simon Krek (Jozef Stefan Institute) Elexis (H2020)
15:00 – 15 min Coffee Break
Posters on the above projects, GOURMET
15-15 – 75 min Poster session
(H2020) and MICE (CEF Telecom)
16:30 End of workshop
Objectives of DG Connect
The overall objective of DG Connect is to contribute to the development of CEF Automated Translation as
a “multilingualism enabler” for CEF DSIs, online services linked to CEF DSIs and other relevant public
online services. This talk will go into detail on the specific objectives.
European Language Grid
This H2020 project will provide a scalable cloud platform, providing, in an easy-to-integrate way, access
to hundreds of commercial and non-commercial language technologies for all European languages.
EU Council Presidency
This CEF Telecom project (“CEF Automated Translation for the EU Council Presidency”) involves the
creation of a toolkit featuring a full set of translation tools for enabling the use of the platform of CEF
Automated Translation in various digital environments, such as the EU Council Presidency websites.
COMPRISE
In this H2020 project (“Cost-effective, Multilingual, Privacy-driven voice-enabled Services”), a fully
private-by-design methodology and tools are developed in order to reduce the cost and increase the
inclusiveness of voice interaction technology.
EMBEDDIA
This H2020 project (“Cross-Lingual Embeddings for Less-represented Languages in European News
Media”) concerns the multilingual future internet and provides new tools allowing high quality
transformations (not translations) between languages.
eTranslation TermBank
This CEF Telecom project provides new terminological resources to the language resource pipeline of the
platform of CEF Automated Translation, for the three sector-specific domains health, business legislation
and customer protection.
NEC TM Data
This CEF Telecom project provides a central pan-European translation memory-sharing platform, where
TMs will be shared directly among public administrations, translation companies and professionals.
APE-QUEST
This CEF Telecom project will boost eTranslation with automated post-editing (APE) and quality
estimation for Electronic Exchange of Social Security Information (EESSI) and Online Dispute Resolution
(ODR).
Prêt-a-LLOD
This H2020 project (“Multilingual Linked Language Data for Knowledge Services across Sectors”) will
reduce efforts in cleaning, organizing and collecting datasets, and combine linked data and language
technologies, using semantic technologies.
ParaCrawl – Corpus Gathering
In this CEF Telecom project, translated sentences are being collected from the web for all 24 official EU
languages plus Icelandic, Norwegian, Basque, Catalan/Valencian, and Galician.
ELRI (European Language Resource Infrastructure)
The focus in the CEF Telecom project lies in the provisioning of an infrastructure to help collect, prepare
and share language resources that can in turn improve translation services
BERGAMOT
This H2020 project (“Browser-based Multilingual Translation”) will add and improve client-side machine
translation in a web browser.
IADAATPA
In this CEF Telecom project (“Intelligent Automated Domain Adapted Automated Translation for Public
Administrations”), a scalable and secure platform is delivered.
ELITR
In this H2020 project (“European Live Translator”), an automatic subtitling system of live meetings and
conference presentations is created. The system is provided with spoken language translation
(interpreting) and automatic minuting.
MARCELL
This CEF Telecom project (“Multilingual Resources for CEF AT in the legal domain”) provides automatic
translation on the body of national legislation in seven countries: Bulgaria, Croatia, Hungary, Poland,
Romania, Slovakia and Slovenia.
Elexis
This H2020 project (“European Lexicographic Infrastructure”) aims at harmonizing European efforts for
the creation of dictionaries by developing tools which reduce the cost and time needed to update
existing or develop new resources.
GOURMET
In this H2020 project (“Global Under-resourced Media Translation”), machine translation systems are
built to translate global news into scarce-resourced languages.
MICE
In this CEF Telecom project (“Middleware for CEF eTranslation”), a middleware layer and dedicated
portal is envisaged for CEF eTranslation for the customisation and improvement of machine translation
output in specific domains.
Katri Tammsaar
Saatja: Luc Meertens <
[email protected]>
Saatmisaeg: reede, 29. märts 2019 12:57
Adressaat:
[email protected]
Teema: [VÄLJAST] Workshops on Language Technology
Manused: Flyer_Customizing Machine Translation.pdf; Flyer_Showcasing Language Technology
in H2020 and CEF Telecom projects.pdf; Workshop #3_agenda.pdf; Workshop #4
_agenda.pdf
Järeltegevuse lipp: Järeltegevus
Olekulipp: Valmis
Dear MICE team,
Please find attached the information on two workshops that we are organising on behalf of DG Connect in Brussels.
These workshops will be held in the centre of Brussels (Hilton Hotel, opposite Brussels Central train station)
April 25: “Customizing Machine Translation” - This workshop gives DSIs and their stakeholders at Member
State level the opportunity to grasp how machine translation (MT) can be deployed in their own
environment and geared towards their specific needs.
April 26: “Showcasing Language Technology in H2020 and CEF Telecom projects” - In this workshop, DSIs
and their stakeholders will be informed about language technology related projects in the H2020
programme and in the CEF Telecom portfolio, as these projects illustrate the potential of context-specific
implementations of language technology.
Attached to this mail you will find a flyer and the detailed programme for each workshop. Workshops are free of
charge, but participants need to register via following link (https://goo.gl/forms/FGKJda3AcNtwISlr2).
Looking forward to meeting you soon,
Best regards,
Luc
Luc Meertens
CROSSLANG
Translation Automation
Amerikagebouw, Kerkstraat 106
9050 Gentbrugge - Belgium
gsm: + 32 477 38 78 30
tel: + 32 9 335 22 00
fax: + 32 9 335 22 01
web: www.crosslang.com
1