Jobs [2 traineeship positions, 5 months]: NLP

[Job posted on January 9th 2015 by The European Commission’s Joint Research Centre (JRC) on Linguist List]

NLP – Terminology

The European Commission’s Joint Research Centre (JRC) is looking to fill
two traineeship positions in the field of:

Terminology discovery over time in the field of disaster risk

If you are interested, please follow the instructions provided at the
URLS listed below. (Code: 2014-IPR-G-000-4154 – ISPRA).

Generic URL:

Job description:

Traineeship rules:


Application deadline:  19 January 2015

Starting date:         around March 2015

Duration:              5 months each

Remuneration:          Up to approximately 1000 Euro per month.

The EMM applications:

JRC-EMM Publications:

Two trainees have been working on this task since September. As the end
users have deemed the results of the motivated and competent team
useful, we plan to continue this work with a new set of two trainees.


The Europe Media Monitor (EMM) group at the European Commission’s Joint
Research Centre (JRC) in Ispra, Italy, is looking for two trainees to
work on a project to automatically explore the development of
terminology in the field of ‘Disaster Risk Management’ (DRM). The
purpose is to give the international stakeholders in that field
(e.g. the United Nations Office for Disaster Risk Reduction UN-ISDR)
concrete and countable evidence of new concepts (terms) emerging in
their field, of changing concepts and of shifts in interest over
time. The study will include both scientific publications and texts
produced by national and international governmental organisations
working in that field.

This first exploratory study will exclusively concern English language
text in the field of Disaster Risk Management, but other languages and
subject areas will be considered in case the outcome of this exploratory
study is deemed concrete and useful. This work may lead to a scientific
publication co-authored by the project contributors.

A scenario to reach this goal of terminology discovery might consist of
the following steps:

(1) Manual or semi-automatic selection and collection of freely
available documents covering the sub-areas of the life cycle of Disaster
Risk Management (Prevention and mitigation; Preparedness; Response;
Recovery and reconstruction);

(2) Conversion of the various file formats (e.g. HTML, PDF, MS-Word)
into a structured text format (e.g. XML);

(3) Selection of suitable off-the-shelf software for the automatic
extraction of terms (e.g. noun phrases);

(4) Usage of this software and, if needed, tuning of this software to
extract lists of potential terms;

(5) Application of statistical methods to select the domain-specific
terms and to weigh or rank them;

(6) Application of statistical methods that allow to observe trends such
as the detection of terms that are more frequently or more rarely used
compared to previous observation periods;

(7) Presentation of the results (term lists, trends) in an
easy-to-understand manner; This may also include a keyword-in-context
presentation of the terms, or similar.

The foreseen traineeship duration is five months, starting around March
2015. The working language in the EMM team is English.


The task is foreseen to be carried out jointly by two trainees who, in
combination, possess the skills or satisfy the criteria listed
below. The combination of a more linguistically inclined person and a
programmer could be fruitful.

– Mature student or post-graduate in any of the following fields (or
similar): computational linguistics, computer science, library
sciences, machine learning;

– Knowledge of – and experience with – freely available Language
Technology tools (e.g. for terminology extraction, term weighting,

– Experience with document format conversion (PDF, HTML, MS-Word etc. to

– Sufficient programming experience to autonomously implement all
necessary steps (Java preferred);

– Knowledge of statistical methods for term weighing (e.g.  chi-square,
TF.IDF) and for automatic categorisation;

– Linguistic sensitivity and an interest for terminology extraction
(what is a term?; relationships between terms);

– Ability to present the project outcome in a format suitable for DRM
specialists who may not be so knowledgeable of Information Technology
(presentation; reporting; visualisation?).

– Ability to work autonomously;

– Team worker;

– Good working knowledge of English plus the ability to communicate in
at least one other official EU language.

In your application, please state your interests and please provide
clear information on your skill set, by elaborating on the
above-mentioned list.  Should you apply as a ready-made team, please
nevertheless clearly state your personal skills and strengths.


The Joint Research Centre (JRC; is the
scientific-technical arm of the European Commission. The approximately
2200 JRC employees working in Ispra are from all EU countries and there
are also some non-EU visitors. The working environment is multilingual,
multi-cultural and multi-disciplinary. The JRC’s Europe Media Monitor
(EMM) team
carries out research and development in the field of text mining
(Language Technology; Computational Linguistics) for the purposes of
media monitoring.  EMM gathers an average of almost 220,000 online news
articles per day in over 70 languages and analyses them to help its
large international user community understand and use this enormous
amount of media information. EMM is publicly accessible via The JRC is also known for having
distributed large quantities of parallel linguistic resources
( , including
JRC-Acquis, DGT-Acquis, JRC-Names, the Translation Memories DGT-TM,
and more.

Ralf Steinberger <> European
Commission – Joint Research Centre (JRC) 21027 Ispra (VA), Italy

URL – Applications:

URL – Resources:

URL – Publications:

Leave a comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s