Given a large population, it is an intensive task to gather individualpreferences over a set of alternatives and arrive at an aggregate or collectivepreference of the population. We show that social network underlying thepopulation can be harnessed to accomplish this task effectively, by samplingpreferences of a small subset of representative nodes. We first develop aFacebook app to create a dataset consisting of preferences of nodes and theunderlying social network, using which, we develop models that capture howpreferences are distributed among nodes in a typical social network. We hencepropose an appropriate objective function for the problem of selecting bestrepresentative nodes. We devise two algorithms, namely, Greedy-min whichprovides a performance guarantee for a wide class of popular voting rules, andGreedy-sum which exhibits excellent performance in practice. We compare theperformance of these proposed algorithms against random-polling and popularcentrality measures, and provide a detailed analysis of the obtained results.Our analysis suggests that selecting representatives using social networkinformation is advantageous for aggregating preferences related to personaltopics (e.g., lifestyle), while random polling with a reasonable sample size isgood enough for aggregating preferences related to social topics (e.g.,government policies).
展开▼
