Experienced researcher position in Statistic and DataMining

Understanding the quality of life of patients with breast cancer:
Medical information extraction and sentiment analysis in health forum
 
An Experienced Researcher position (e.g. post-doc, 10 months), funded by
the GIS-IRESP (http://www.iresp.net/), is available in Montpellier
(France) at the LIRMM laboratory (Laboratory of computer sciences,
robotic, micro-electronic of Montpellier, University of Montpellier,
CNRS, http://www.lirmm.fr/), the I3M Laboratory (Institute of
Mathematics and mathematical modelling of Montpellier,
http://ens.math.univ-montp2.fr/) and the ICM (Institute of Cancer
Montpellier, http://www.icm.unicancer.fr/).

Topic: The proposed topic is part of a larger project called "Comparison
of longitudinal analysis models of the health-related quality of life
(QoL) in oncology". The overall aim of this project is to propose and
compare QoL longitudinal analysis methods based on a methodology already
used in survival analysis or on more complex theoretical models (linear
mixed model, IRT model). These methods will be applied to clinical trial
data to analyse the EORTC QLQ-C30 questionnaire in a variety of
therapeutic situations and tumor sites. The sentiment analysis extracted
from health forum will appear to be an interesting alternative to
quality of life study based on standardised auto-questionnaires and will
provide a perspective via results coming from the QLQ-C30 analysis. This
project will be focus on women with breast cancer and on the development
process of the disease to take into account the longitudinal
aspect. Eight researchers have been working on this topic since early
2013. We are now looking for a researcher who will work together with
this team. The originality of this post-doc is to be part of a
multidisciplinary project and to develop semi-automatic medical
information extraction methods and sentiment analysis methods applied to
health forum. The main objective is to provide healthcare professionals
and patients with a knowledge discovery framework.

The extraction of relevant information within health forum is a first
step to define a semi-automatic method for sentiment analysis related to
cancer management (e.g., identification of latent topics that describe a
corpus of messages from several health forums). One of the main issues
is to propose original methods taking into account the heterogeneity of
authors, topics, texts in their structure and their contents. This
research work focuses initially on the automatic extraction of features
such as medical, opinion, risk... features. In a second step, the research
will develop useful views to present these features to the targeted
users: health care professionals and patients. In this highly subjective
context, the characterization and understanding of the representations
of the patients medical concepts is difficult. However, these tasks are
particularly fundamental to plan and design medical public
policies. Finally, the work proposed in this post-doc should help meet
this dual challenge: medical information extraction and sentiment
analysis related to cancer management.

Under the overall guidance of the research coordinator at LIRMM, I3M and
ICM, the researcher will also contribute to the preparation of project
deliverables, publish related work in scientific journals, participate
in workshops and meetings and track schedules of tasks. The position is
a research-only post, with no teaching obligations.

Location: The work will be carried out at the LIRMM laboratory
(Montpellier in France).

The consortium of the project is composed of computer science
researchers, statisticians as well as health care professionals.

Supervision: The project will be formally jointly supervised by
Dr. Sandra Bringay at LIRMM, Dr. Cyrille Joutard and Pr. Christian
Lavergne at I3M and Dr. Mollevi Caroline at ICM.

Requirements: The candidate is required to hold a PhD obtained in a
topic related to text mining, statistic or automatic natural language
processing, with a firm background in sentiment analysis and focus on
the development and implementation of web-based tools. Candidates with
backgrounds in medical datasets are strongly encouraged to apply. Strong
experience in programming language (e.g. C++, PERL, PHP...) is
requested. Working experience at national and international levels is an
asset. The candidate will work in a team with PhD researchers and MSc
students. The candidate should demonstrate team building skills, and
ability to work collaboratively and to establish and maintain effective
working relationships with colleagues of different nationalities and
cultural backgrounds. The candidate should prove ability to analyse and
integrate diverse information and to communicate it (verbally and in
writing) clearly and effectively. For the advertised post, fluency in
oral and written English is required. Knowledge in French is an asset.

It is expected that the successful candidate will participate in all
meetings and workshops organized by the project - related additional
funding will be available. The candidate must be willing to disseminate
the results within the project during the meetings, but also and in
particular through presentations at conferences and contributions to
scientific journals.

Starting date: October 2013 (some flexibility is possible, but at the
latest January 2014).

Salary: 1850¤ (Net Income)

Application deadline: 15th July 2013, applications will be processed as
they are received.

Contact: Interested candidates should contact and send their application
(curriculum vitae, statement of research interest, publications, and
addresses of three referees) by email to:

 
Dr. Sandra BRINGAY
MCF LIRMM - UMIII
bringay@lirmm.fr
LIRMM, UMR 5506
161 rue Ada
34095 Montpellier Cedex 5
Tél : (33/0) 04 67 41 86 31