Social Choice in a Social Context: A Multivariate Algorithmics Perspective

该研究项目的目标是修改和扩展已建立的计算社会选择模型，以捕捉由代理人（例如选举中的选民）的社会邻域引起的影响。这种影响可能是合作的味道（例如，代理人可以合作，如果他们知道对方），竞争的味道（例如，代理人的行为策略和预期他们的对手也有策略），或基于隐私问题（例如代理人不喜欢透露他们的完整偏好）。我建议分析这种影响的算法的复杂性。为此，将使用算法博弈论，多元算法，社交网络和投票理论的概念。为了确定易于处理的案件，受限制的偏好领域将发挥决定性作用。

The goal of this research project is to revise and extend established models of Computational Social Choice in order to capture effects induced by the social neighborhood of the agents (e.g. voters in an election). Such effects may be of cooperative flavor (e.g. agents may cooperate if they know each other), of competitive flavor (e.g. agents behave strategically and anticipate their opponents to also behave strategically), or based on privacy issues (e.g. agents prefer not to reveal their complete preferences). I propose to analyze the influence of such effects on the algorithmic complexity. To this end, concepts from algorithmic game theory, multivariate algorithmics, social networks, and voting theory will be used. For the identification of tractable cases, restricted domains of preferences will play a decisive role.