Eigenvalue-based Bipartite Ranking
Stephan Link · 2011
Rankings have a high relevance in many different domains, such as science, sports or education. According to the needs of applications and their data several ranking techniques were developed. In this thesis, we introduce a ranking technique called bipartite ranking, that is based on PageRank [8]. Bipartite ranking is suitable to rank scenarios, that can be represented as a bipartite graph what is among others the case for Rankings with feedback ranking in which those who were rated rate the person who rated them. We show, that the ranking results of bipartite ranking have a higher objectiveness towards defined criteria than existing ranking techniques. And we give examples for real situations, where bipartite ranking could be used in.