Job: Fellowship in WIPO in the field of Statistical Machine Translation (4 months)

[Thierry Hamon]

Date: Fri, 8 Apr 2011 17:03:43 +0200
From: "Pouliquen, Bruno" <Bruno.Pouliquen at wipo.int>
Message-ID: <9EA20A207BA4FB409FD9161056B20AB30BCA212E76 at ICCV101C.gms02.unicc.org>
X-url: http://www.statmt.org/moses/
X-url: http://www.wipo.int/
X-url: http://www.wipo.int/patentscope
X-url: http://www.wipo.int/patentscope/translate


Fellowship in WIPO in the field of Statistical Machine Translation (4 months)

We are looking for a skilled computer scientist to assist our work in
Statistical Machine Translation at the World Intellectual Property
Organization in Geneva (Switzerland).

Project:
The WIPO Global Databases Section is working in the area of
automatically translating Patent documents using open source Moses
(http://www.statmt.org/moses/) and is especially looking for a
candidate that could contribute in one or more of the following
fields:

- Adapting Moses core software to allow dynamic parameterization
  during decoding (good knowledge of C++ required). Includes dynamic
  setting of language models

- Pre and Post processing texts in different languages to improve
  translation quality (including, but not limited to: Chinese, Korean,
  Japanese and German languages)

- Experimenting with different settings of Moses (tree-based models,
  use of dynamic suffix arrays, etc.)

- Triangulation: improving SMT through the use of another pivot
  language (e.g. improving English->German through the use of
  English->French and French->German).

- Graphical User Interface (good knowledge of Java/Swing and/or
  Web/JSF 2.0 required) to provide an improved means for accessing
  various proposals for translations

Our Organization:

The World Intellectual Property Organization (WIPO, see
http://www.wipo.int/) is a specialized agency of the United
Nations. It is dedicated to developing a balanced and accessible
international Intellectual Property (IP) system. As part of its
mandate, WIPO translates Patent applications and disseminates
information about published patent applications using PATENSCOPE
search engine: http://www.wipo.int/patentscope. To make this
information available world wide WIPO is looking for techniques to
help the translation of patents in various languages. A first
prototype for patent translation is available at:
http://www.wipo.int/patentscope/translate

Required skills:

- The candidate must have a strong background in computer science,
  ideally with specialization in computational linguistics, and
  preferably familiar with Moses (or other SMT tools).

- Depending on the field, a strong knowledge of C++ and/or Java is
  required

- Applicants should have excellent written and spoken English or
  French, a working knowledge of other official languages of WIPO
  (German, Spanish, Portuguese, Russian, Arabic, Chinese, Japanese or
  Korean) would be an advantage. We particularly welcome written
  knowledge of Chinese, Japanese or Korean.

- Good knowledge of web technology, databases, statistics and search
  engines (Lucene / Solr), scripting languages (Perl, bash) would also
  be an advantage.


Deliverable:
After four months the candidate should have developed a tool which
improves the SMT process in WIPO.
He/she should have also written a report on his tool, ideally this
report could be submitted to a scientific conference (or journal) in
the field (with candidate as first author).


Additional information:
Remuneration: 4000 to 6000 CHF per month depending on qualifications
(travel and accommodation costs to be borne by the applicant).
Expected starting date: 1st of June 2011.
Duration: 4 months, please note that this is a short term position

Please send your questions and application with detailed CV by email
to: patentscope at wipo.int

-----
Bruno Pouliquen
Global database section
WIPO

World Intellectual Property Organization Disclaimer:

-------------------------------------------------------------------------
Message diffuse par la liste Langage Naturel <LN at cines.fr>
Informations, abonnement : http://www.atala.org/article.php3?id_article=48
English version       : 
Archives                 : http://listserv.linguistlist.org/archives/ln.html
                                http://liste.cines.fr/info/ln

La liste LN est parrainee par l'ATALA (Association pour le Traitement
Automatique des Langues)
Information et adhesion  : http://www.atala.org/
-------------------------------------------------------------------------