Job: Stage de recherche, Technicolor (Rennes), Analysis of Web forums

[Thierry Hamon]

Date: Tue, 16 Oct 2012 16:56:58 +0200
From: Guegan Marie <Marie.Guegan at technicolor.com>
Message-ID: <FABFECD6AE260643A25A5C6F611BCC0F19ECDD4961 at MOPESMBX01.eu.thmulti.com>
X-url: http://www.technicolor.com/

Internship position available at Technicolor R&D in Rennes.

Title
------
“Which scene are you talking about?”
Recognizing scenes discussed on cinema or TV forums.

Context
-------
For more info on Technicolor Research & Innovation, Rennes :
https://research.technicolor.com/rennes/

The internship will be hosted at Technicolor R&I in Rennes, France (500
employees, of which 130 researchers), within the Media Computing
Lab. Our lab aims at bringing modern trends in computing to the service
of novel media engines in content creation (visual effects, animation)
as well as content discovery and retrieval.
More specifically, our team focuses on Web user comments posted on
forums and social networks. We use various approaches such as data
mining, social network analysis and natural language processing.

Objective
---------
This internship aims at designing, developing and evaluating an
information extraction system for user comments. The domain is dedicated
to cinema and television. Each comment is already attached to a
particular audiovisual content (movie, TV series, TV program). One of
our goals is to detect within the comments the text segments which refer
to a particular moment in the video. For instance, users may talk about
their favorite scene or quote a famous dialogue.

Task description
-----------------
We will not analyze the audiovisual signal (image or audio), but solely
the text of comments. Comments have already been collected and saved
into a database. Hence the internship will focus on the analysis of the
dataset rather than its collection. The intern will be responsible for
choosing best techniques, based on a survey he or she will conduct on
state-of-the-art approaches.
The developed system will be evaluated by the intern, both
quantitatively and qualitatively. Depending on obtained results and
innovative ideas, this work may lead to a research publication in a
conference.

Keywords
---------
Natural language processing (NLP), machine learning, data mining, text
mining

Profile of the candidate
-------------------------
* Student in final year of master or science engineering school
* Computer science (Python, Java)
* Skills in machine learning, data mining and natural language
  processing
* Strong interest in research (will constitute a survey)
* Interest in social networks
* English mandatory
* Appreciates working with a team spirit

Internship period & duration
-----------------------------------
6 months, starting preferably around February or March 2013, depending
on the candidate’s constraints.

Please email your CV and cover letter to
stage.rennes at technicolor.com<mailto:stage.rennes at technicolor.com> with
reference [TRDF-DM-029] in the subject.


Marie Guégan

Media Computing Lab
Research & Innovation
Technicolor R&D France
975 avenue des Champs Blancs, CS 17616
35576 Cesson-Sévigné Cedex, France
www.technicolor.com

Important : Technicolor R&D France déménage.
Notre nouvelle adresse, à compter du 24 octobre 2012 devient :
975 avenue des champs blancs, CS 17616, 35576 Cesson Sévigné, France -
tél (standard) : +33 (0)2 99 27 30 00 (inchangé).

-------------------------------------------------------------------------
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/
-------------------------------------------------------------------------