univ Le Havre)
geraldine.del_mondo@insa-rouen.fr
eric.sanlaville@univ-lehavre.fr
*Team Coordination Dynamics : a team seen as an adaptive complex system,
modelized by a temporal semantic graph*
This project, funded by the ANR for a 3-year period starting January 1,
2024, brings together specialists in computer science and in cognitive
science applied to competitive sports. The main objective of DynaTeam is to
propose graph-theoretic models for identifying and then predicting team
strategies in various team sports (basketball, rugby, handball, etc.). The
proposed models must take into account the temporal and spatial aspects of
interactions between players, both in matches and during training sessions.
In the second type of situation, the way in which the coach's instructions
are taken into account will be analyzed.
The DynaTeam teams are already collaborating on these modeling aspects in
various projects. So far, the models proposed take into account the
networks of passes between players during the possessions of a match thanks
to dynamic graphs that ignore spatial aspects [1]. A major objective of the
project is to propose and validate models that take greater account of all
aspects (temporal, spatial, coach's instructions).
Example of recent works on the subject by the team (available on hal.fr) :
[1] Bourgeais, Q., Hacques, G., Charrier, R., Sanlaville, E., & Seifert, L.
(2023, July). Identify team sport team signature through passing network.
In *European College of Sport Science*.
[2] Oberoi, K. S., & Del Mondo, G. (2021, May). Graph-based pattern
detection in spatio-temporal phenomena. In *16th Spatial analysis and
geomatics conference (SAGEO 2021)*.
*Activity details*
The person recruited will participate in Work Package 1 (analysis tools),
and more specifically in Task 1.2: Dynamic graph integrating the spatial
characteristics of players on the playing field.
The candidate will propose new, more general models that take into account
both temporal and spatial aspects. He/she will draw on the two research
themes of temporal graphs and semantic graphs. The proposed models will be
tested using field data from matches and training sessions. They will be
accompanied by indicators to differentiate the behavior of different teams
under different real-life conditions.
The candidate will work closely with the 3 teams and the other people
recruited for the project, including a post-doc creating a prediction model
using graph-based machine learning, and another post-doc studying the
emergence of adaptive collective behavior according to training and
prescription conditions.
*Skills and abilities required*
The candidate will hold a PhD in Computer Science or Applied Mathematics.
He or she must have in-depth skills in graph algorithms, and be able to
implement, validate and test the models and algorithms designed. Knowledge
of temporal graphs (or dynamic graphs) and/or semantic graphs is a plus, as
is initial experience in social network applications, particularly in
sports.
*Practical conditions*
The candidate will be under contract to the University of Le Havre. He/she
will work within the RI2C team at LITIS in Le Havre, but in constant
contact with the APP team at LITIS in Rouen, and in particular with
Géraldine Del Mondo (and Sébastien Adam). Frequent travel to the Madrillet
campus in Rouen is required, and will be covered by the project budget. The
post doc is for 12 months, renewable once.
*Contact* :
Géraldine Del Mondo (LITIS INSA ROUEN) & Eric Sanlaville (LITIS univ Le
Havre)
Geraldine.del_mondo@insa-rouen.fr
Eric.Sanlaville@univ-lehavre.fr
**********************************************************
*
* Contributions to be spread via DMANET are submitted to
*
* DMANET@zpr.uni-koeln.de
*
* Replies to a message carried on DMANET should NOT be
* addressed to DMANET but to the original sender. The
* original sender, however, is invited to prepare an
* update of the replies received and to communicate it
* via DMANET.
*
* DISCRETE MATHEMATICS AND ALGORITHMS NETWORK (DMANET)
* http://www.zaik.uni-koeln.de/AFS/publications/dmanet/
*
**********************************************************
Posts
