A Reputation-Based Model for Decision-Making in Online Social Networks

TL;DR

This paper introduces a reputation-based centrality measure for online social networks, demonstrating that strategic reputation maximization leads to realistic social behaviors and robust network structures.

Contribution

It proposes a novel eigenvector-like centrality measure and shows its effectiveness in modeling decision-making and reputation dynamics in online social platforms.

Findings

Maximizing reputation leads to stable 'like' exchange rates.

Networks with real-world features support reputation-seeking behavior.

The model reflects realistic social dynamics in online communities.

Abstract

The online exchange of social recognition including, for instance, the Facebook "like" appears to produce a scarce allocation without a clear utility function defined for anyone involved. Given the importance attached to such digital commodities by both users and advertisers, it is of interest to study the forces governing their economics. Here we propose a centrality measure akin to eigenvector centrality to describe an individual's perceived importance in an online social network. It is shown in silico that strategically maximizing this prestige metric results in finite nontrivial rates of "like" endowment. Furthermore, it is found that systems of reputation-seeking agents are supported most robustly by networks with the features of real human societies including preferential attachment and the small-world property. We conclude that the incentive system studied here can produce realistic behavior and may therefore provide a framework for a more general model of decision-making in online communities.