Optimized Directional Content Distribution Using Reputation

Abstract

Word of mouth represents, from historical times, a powerful mechanism employed within human societies for influencing the behavior of their members. Translated into the computational world, the same feedback mechanism preserves or even broadens its impact. This paper presents an intriguing perspective on using reputation systems: reputation is gained by complying with the norms and norms are modified by the agents with high reputation. The aim of the model proposed by this paper is to optimize content distribution within a system. The model achieves this goal by prioritizing the preferences of the highest reputed agents. The reputation of an agent is translated into the influence it gains in altering system norms. The computational reputation model proposed requires minimum resources being thus well suited for agents running on mobile devices. Being completely decentralized, the interaction architecture provided by the model offers a high degree of scalability.