An Empirical Study of Voting Rules and Manipulation with Large Datasets
The study of voting systems often takes place in the theoretical domain due to a lack of large samples of sincere, strictly ordered voting data. We derive several million elections (more than all the existing studies combined) from a publicly available data, the Net´flix Prize dataset. The Net´flix data is derived from millions of Net´flix users, who have an incentive to report sincere preferences, unlike random survey takers.
We use these elections to investigate theoretically-possible phenomena, including different voting rules yielding different winners, and the possibility of irrational aggregated preferences. We investigate the occurrence of statistical patterns of interest, such as single- peakedness. In addition, we look at the practical manipulability of elections according to several recently-proposed manipulation algorithms.
This is joint work with Nicholas Mattei and James Forshee, and has appeared in Dr. Mattei's dissertation and in a paper at ComSoc 2012.
Dr. Judy Goldsmith got her AB from Princeton University and her MS and PhD from the University of Wisconsin-Madison, in mathematics. She has worked on computational complexity, artificial intelligence, and comparative decision making studies. She was honored by the AAAS in 1998 for her work in mentoring women and other members of underrepresented groups in STEM fields.
Dr. Goldsmith is on the editorial board for the Journal of AI Research, and has guest-edited special issues of AI Magazine, IJAR, AMAI, and others. She has co-organized many workshops and conferences, most recently a special session on computational social choice at ISAIM 2012, a 2011 conference on Comparative decision making, the IJCAI 2011 doctoral consortium, and the 2012 Midwest Morris Ale (a dance event).
In the 2012-13 academic year, she is teaching Introduction to AI; Comparative Decision Making Studies; and Science Fiction and Computer Ethics.