On the subject there are many interesting discussions and theories in human social psychology and philosophy, but without models or algorithms. My question relates specifically to the multi-agents in technique, computer science or artificial intelligence (eg web or robot agents), and other categories. In general, the known algorithms for learning are for individuals (ANN, etc.), not for assemblies of individuals. There are updates or new approaches in this area?
Dear Eugen,
Any ABM in which agents are being selected by their environment results in population level Learning. Evolutionary models in general are models of group learning. Some approaches combine population level learning through selection with exchange of information between agents, leading to faster group learning than with selection alone.
Hello Eugen
The Journal of Artificial Societies and Social Simulation (JASSS) is a great source of information regarding this topic!
http://jasss.soc.surrey.ac.uk/JASSS.html
http://jasss.soc.surrey.ac.uk/admin/searchresults.html?q=group%20dynamics
Hello, in our team we work on the subject and our contribution consist in the analysis (by a MAS) of textual messages between students in order to distinguish sociological (learner) profiles. For the moment the publications linked to the subject are in french but we are working on an english version. I can answer to questions if you are interested.
Conception et réalisation d’un système intelligent pour l’analyse automatique de conversations textuelles asynchrones d’apprenants pour la détermination de comportements sociaux
Youness CHAABI, Messoussi ROCHDI, Vincent HILAIRE, Khadija LEKDIOUI, Yassine RUICHEK, and Raja TOUAHNI.
In Proc. of SITA'14 - 9th International Conference on Intelligent Systems: Theories and Applications, 2014.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence