Equilibrium-Based Voting: A Strategy for Electing Service Providers in P2P E-Learning

Filtering trusted learning resources is paramount for a successful P2P e-learning future. Unlike centralized systems which feature a centralized authority that maintains reputation values of all users in the system, in decentralized P2P networks, the popular approach is using a decentralized mechanism to achieve global estimates of peers' reputations. While decentralized reputation systems solve part of the problem, there is still the well-known problem of manipulating reputation systems and voting mechanisms to achieve selfish objectives or satisfy malicious intent. In the context of P2P e-learning, interacting with untrustworthy service providers, i.e. manipulative or cheating nodes, may suppress the e-learning process. We approach this problem by presenting an upgrade of reputation-based voting mechanisms, called equilibrium-based voting (EquiVote).We show by empirical analysis that our approach achieves verification and correctness of erroneous reputation values, caused by preelection manipulation activities, at the time of electing learning service providers in P2P e-learning environments.