Computational Social Choice in the Clouds

Abstract

In the era of big data we are concerned with solving computational problems on huge datasets. To handle huge datasets in cloud systems dedicated programming frameworks are used, among which MapReduce is the most widely employed. It is an important issue in many application areas to design parallel algorithms which can be executed efficiently on cloud systems and can cope with big data. In computational social choice we are concerned with computational questions of joint decision making based on preference data. The question of how to handle huge preference datasets has not yet received much attention. In this report we summarize our recent work on designing and evaluating algorithms for winner determination in huge elections using the MapReduce framework.