*****  To join INSNA, visit  *****

A PhD position is available at the LIA (computer science laboratory of the
University of Avignon, France), in collaboration with the software company
Nectar de Code (, starting from October 2015
and cofunded by the administrative PACA region and Nectar de Code.

*Title: *User models for the automatic monitoring of social networks

*Keywords: *machine learning, modeling, natural language processing,
networka analysis

*Description: *
(in French, contact us for more details in English)

*Laboratory: *LIA (

*Advisors: *Georges Linarès ([log in to unmask]), Richard Dufour
([log in to unmask]), Vincent Labatut (
[log in to unmask])

*Funding: *Provence region (public structure), Nectar de Code (software

*Conditions: *the candidate must be less than 30 years old

*Starting date:* October 2015

*Contacts: *[log in to unmask], [log in to unmask]
, [log in to unmask]

*Required files: *send the following files to all three advisors: CV +
*topic-related* motivation letter + grade transcripts for both Master years
(or equivalent)

*Application deadline:* 9th of june 2015 (*note* *the very short delay,
beyond our control*)

*Summary: *

In the recent years, the interactive aspects of the Internet became
increasingly important, transforming it into the main support for free
communication between potentially very different persons. This led to
considerable scientific and societal issues related to the supervision of
Web-based social interactions. Indeed, in order to work correctly, open
communication spaces need to be monitored, which is a difficult task.
Without any moderation, such a service can be hijacked, or end up hosting
negative behaviors decreasing the service quality, or even illegal
activities. One could argue the absence of any control actually contributed
to the development of such open spaces on the Web, however the absence of
any moderation could also dramatically limit their interest. Human-operated
moderation is often costly, and even economically unfeasible; creating
automatic monitoring methods for such large-scale textual interactions is
consequently of the highest interest. We propose to develop an approach
focused on users, and to evaluate it on the specific case of a social
networking service. Tackling this problem involves designing models of the
users’ behavior, and defining methods to estimate these statistical models.
The general framework of this PhD topic is social networks analysis.

Best regards,
vincent labatut
associate professor
university of avignon - computer science lab <>
 / teaching center <>
[log in to unmask] - academia page
+33 (0)4 90 84 35 25 - room ET2

SOCNET is a service of INSNA, the professional association for social
network researchers ( To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.