PowerPoint-Pr

1
Thematic Network ENBI European Network for
Biodiversity Information Objective To manage an
open network of relevant biodiversity information
centres in Europe and other countries of the
western European palearctic region. Task IfM To
provide multi-lingual access to biodiversity
information in the Internet The work plan is
focused on user needs, and on making European
biodiversity information available for the
end-users. The users include government agencies,
decision makers, legislators, scientists,
companies, and citizens. Also non-European users
are very dependent on access to European
information, because many data in European
repositories originate from non-European (often
developing) countries. Understanding the needs of
all these kinds of users is paramount for the
dissemination of biodiversity knowledge
resources, and common access, with attention to
multilingual access (WP 11), is a key issue
2
Thematic Network ENBI European Network for
Biodiversity Information ENBI's main objective is
to establish a strong network that will identify
biodiversity information priorities to be managed
at the European scale. ENBI is also the European
contribution to the Global Biodiversity
Information Facility (GBIF). ENBI is an EC
supported Thematic Network accommodating 65
European institutes representing 24 countries.
The project started on the first of January 2003
and will run for 3 years. Total project budget
3 Mill. Euro Task IfM WP 11, Multi-lingual
access for European biodiversity
site Contribution to IfM (Coordination of WP 11)
250,000 Euro
3
What is GBIF? The GBIF Mission The purpose of
the Global Biodiversity Information Facility
(GBIF) is to make the world's biodiversity data
freely and universally available. GBIF works
cooperatively with and in support of several
other international organizations concerned with
biodiversity. These include the Clearing House
Mechanism and the Global Taxonomic Initiative of
the Convention on Biological Diversity , and
regional biodiversity information networks.
Participants in GBIF have signed the Memorandum
of Understanding, and support network Nodes
through which they provide data. The GBIF
Vision GBIF contributes to economic growth,
ecological sustainability, social outcomes and
scientific research by increasing the utility,
availability and completeness of primary
scientific biodiversity information available on
the Internet. Convention on Biological Diversity
Rio, 1992 "Clearing-house mechanism" to ensure
that all governments have access to the
information and technologies they need for their
work on biodiversity.
4
WP 11. Multi-lingual access for European
biodiversity site Start date or starting
event March 2003 Name of the partner
responsible Institute of Marine Research at the
University of Kiel, Co-ordinator WP 11, Bernd
Ueberschär N of the partner responsible P11 Ns
of other partners involved P1, P8, P15, P21,
P31, P42, P48, P51, P62 Person-months per
partner (P11 24)

• Objectives
• Identify biodiversity terms (biology, morphology,
  taxonomy, geography, genetics) that need special
• dictionaries for proper translation, as indicated
  by unsatisfactory translation by the existing
  service
• Translate these terms into 8 European languages
  in a format that can be used efficiently for
  machine
• Translation
• Explore options to build a prototype biodiversity
  glossary that can be used by European
  biodiversity web
• Sites
• Explore options to significantly improve access
  to biodiversity information through vernacular
  names
• Define recommendations on how to tackle these
  issues in subsequent dedicated work packages,
  including finding a long-term host for the
  biodiversity translation service

5

• Introduction WP11
• 80 of the Internet's content is in English43
  of Internet users today cannot read English at
  all
• Multi-lingual access will be provided to European
  biodiversity sites through a user-friendly
  interface on the World Wide Web. (Dutch, French,
  German, Greek, Italian, Portuguese, and Spanish).
  
• Traditional 'manual' translation is not an
  option. Rather, machine translation on demand has
  to be applied. The quality of machine translation
  varies greatly. Results can be drastically
  improved if specialized dictionaries are
  available for the topic in question, and certain
  terms are excluded from translation.
• The Translation Service of the European
  Commission (SDT) has developed its own machine
  translation system, starting in the 1970s and
  building on the Systran engine, currently
  supporting 8 European languages and 18 language
  pairs. The service provides machine translation
  of documents for registered users.
• SDT plans to add machine translation of web pages
  to its services in 2002. This work package will
  create special biodiversity dictionaries to be
  integrated in the machine translation service of
  the European Commission. It will improve access
  to biodiversity information through vernacular
  instead of scientific names. It will provide a
  glossary explaining unfamiliar biodiversity terms
  in 8 European languages.
• European biodiversity web sites can avail of this
  service by showing a 'Translate' button on their
  pages. The activities in this work package will
  provide vernacular names to the GBIF Electronic
  Catalogue of Names.

6
Multi-lingual access for European biodiversity
sites through a user-friendly interface on the
World Wide Web is the main issue of WP 11
About Machine translation
The principle of machine translation a raw
translation of a document, from a source language
into a target language, is made on the basis of a
system of dictionaries and linguistic programs
(e.g. SYSTRAN).
Machine translation quality
Machine translation quality mainly depends on the
kind of documents (with typing errors or complex
syntax, the result will be poor) language
similarities and on specific dictionaries
available.
7
Systran offers the service SYSTRANLinks
SYSTRANLinksSYSTRANLinks is a turnkey website
translation solution. SYSTRANLinks transforms
standard websites and content applications into
interactive multilingual hubs, all within
seconds. SYSTRANLinks offers all major European,
Asian and Russian languages .
8
The European Commission offers EC SYSTRAN
The Commission has been developing the Systran
machine translation system since 1976. EC-SYSTRAN
has been developed for internal purposes and is
therefore distinct from the commercial version
available and on Web sites such as Altavista. The
system can produce 2 000 pages of raw translation
per hour. Machine translation is accessible from
a PC via a web interface and the electronic mail
system. The translation is returned to the user
within minutes.
On request, EC SYSTRAN can be made available to
public authorities, schools and universities in
EU Member States (permission granted already for
the life time of this project)
9
Customizable Dictionaries
SYSTRAN Dictionary Manager guides through the
process of adding own terms and expressions.
SYSTRAN Dictionary Manager even allows to import
and export Text and Excel file-formatted
glossaries, and to create your own custom domains
for greater term specification.
10
About Eurodicautom (the Translation Service's
terminology bank)
Eurodicautom is the European Commissions
multilingual term bank. It an invaluable tool for
translators, interpreters, terminologists and
other linguists worldwide over the
Internet. Eurodicautom covers a broad spectrum of
human knowledge, but is particularly rich in
technical and specialised terminology
(agriculture, telecommunications, transport,
legislation, finance) related to EU policy. New
data are added constantly, obtained from
Commission terminologists, translators, linguists
from other European and international
institutions, research centres, publishers,
private experts, etc. Entries are classified into
48 subject fields (ranging from medicine to
public administration). At present the term bank
contains about five and a half million entries
(terms and abbreviations), subdivided into more
than 800 collections.
Web Interface of Eurodicautom
11
Return Page of Eurodicautom (example request for
"finfish)
Eurodicautom will help to translate Biodiversity
terms. In turn, the project will add new data to
the system
12
Time Table WP11, Multilingual Access
'How to proceed' workshop conducted. Report with
documentation on how to further develop
multi-lingual access to biodiversity
W
Proof of concept of vernacular names to GBIF
Recommendations on how to access to biodiversity
information through common names available
Beta version of translation service (M18)Online
service for quality translation of biodiversity
web pages (M 26). Test with FishBase
Integration of dictionaries with SDT machine
translation. Connect, test, improve machine
translation
Prototype Biodiversity Glossary available
Development prototype of biodiversity glossary
Author contracts for all languages in place
Negotiation Author Contracts
Workshop on how to use and to translate
biodiversity dictionaries conducted, Report
W
Options for providing vernacular names to the
GBIF Electronic Catalogue
Options explored, Report
Identification of biodiversity terms that need
special dictionaries for proper translation
List of terms available in English, translation
into German available
Tasks
Usefulness of existing public-domain glossaries
Glossaries identified, permissions obtained
Deliverables
Months
13
Provisional Summary of Tasks "Multilingual
Access", WP11

• Explore the (technical) options and abilities of
  the EC SYSTRAN system, e.g. how to compile
  specialized dictionaries, how to provide
  dictionaries to the EC SYSTRAN system, how to
  proceed with Website "on-the-fly translation
  (date with the Translation Service arranged for
  tuesday afternoon).
• Compile a list of biodiversity terms (biology,
  morphology, taxonomy, geography, genetics) that
  need special dictionaries for proper translation,
  under consideration of the needs an input of
  ENBI-Partners and applying the Eurodicautom
  system.
• Compile dictionaries of those terms in 8
  languages (with European partners).
• Application of the EC SYSTRAN system, test with
  the FishBase Website.
• Make the technique available for other
  biodiversity databases
• Explore options to significantly improve access
  to biodiversity information through vernacular
  names
• Results of the workshop Time table in place,
  Partner from France, Greece, German for
  specialized dictionaries/glossaries in place,
  still open from Spain, Portugal, Italy, Dutch.
  Launching a website (www.multilinguaweb.org).
  First Workshop date fixed (beginning of October).
  Agreement with EC Translation Service in place
  (explore the feasability of special needs for
  this project, e.g. as Web Site Translation "on
  the fly" encoding).