Seminar

In this talk I will discuss how numerical experiments on elections were typically performed in computational social choice and what we can learn from it. In particular, we will see what election sizes were considered that statistical cultures were used etc. Further, we will analyze which choices were reasonable and under what conditions. On the one hand, the talk is based on the "Guide to Experiments" project, which collects experimental election papers from AAAI/IJCAI/AAMAS conferences held since 2010. On the other one, we will describe and use the map of elections framework. This framework is based on the idea that one can represent a dataset of elections as a set of points on a plane, so that the closer two elections are to each other, the more similar they are. Such maps can be used, e.g., to compare statistical cultures to each other, and to compare them to real-life elections.