| KEYWORD |
Social network analysis applied to football matches using video analysis
Thesis in external company
keywords IMAGE PROCESSING, MACHINE LEARNING, ARTIFICIAL NEURAL NETWORKS, SOCIAL NETWORK ANALYSIS
Reference persons MAURIZIO MORISIO
Research Groups DAUIN - GR-16 - SOFTWARE ENGINEERING GROUP - SOFTENG
Thesis type EXPERIMENTAL
Description The specific objective of this project is to implement a tool capable of tracking a set of players in
order to be able to analyze data at an individual and team level. Examples of data to analyze could be
passes, throws, the direction of sprints, etc.
Techniques and methods based on the analysis of social networks can be used to analyze the
dynamics of connection between the members of a team and their opponents.
Social network analysis in team sports is commonly used to analyze the passes performed between
teammates. Nevertheless, other linkage indicators such as defensive marking, defensive coverage, or
attacking coverage can be also analyzed.
The codification of players can be made by considering the tactical position or using the specific
code of a player. The first alternative provides a better analysis of the patterns of play and collective
organization of the team. On other hand, the second alternative provides information about the
specific contribution of a player to the team’s network;
The adjacency matrices that come from the observational processes can be used in different SNA
software to process macro-, meso- and microanalysis.
The project includes data collection and the identification of appropriate tools (e.g. Python, OpenCV,
Tensorflow, Ucinet, Pajek, PATO).
The thesis is expected to last approximately 6 months. The student will work within the Orbyta
analytics team.
Required skills Python (OpenCV, TensorFlow), Machine learning, image processing
Deadline 27/10/2022
PROPONI LA TUA CANDIDATURA