The Computational Difficulty of Bribery in Qualitative Coalitional Games

Printer-friendly version
Working paper
Author/s: 
Andrew Dowell, Michael Wooldridge, Peter Mcburney
Issue number: 
2007.100
Publisher: 
FEEM
Year: 
2007
Qualitative coalitional games (QCG) are representations of coalitional games in which self interested agents, each with their own individual goals, group together in order to achieve a set of goals which satisfy all the agents within that group. In such a representation, it is the strategy of the agents to find the best coalition to join. Previous work into QCGs has investigated the computational complexity of determining which is the best coalition to join. We plan to expand on this work by investigating the computational complexity of computing agent power in QCGs as well as by showing that insincere strategies, particularly bribery, are possible when the envy-freeness assumption is removed but that it is computationally difficult to identify the best agents to bribe.
Developed by Paolo Gittoi