Insights about Fairness of Content Distribution in Online Communities
Online communities and business communities are often characterized by their unequal distribution of interactions with targets such as content and users: Few users are interacting with many targets, while many users are attached to only few targets. The inequality of such network structures is typically measured using the power-law exponent. However, this approach that is frequently used in network studies has several weaknesses, such as its narrow applicability and expensive computational complexity. Beyond the fact that power laws are by far not a universal phenomenon in online networks, the power-law exponent has the surprising property of being negatively correlated with the usual notion of inequality, making it unintuitive as a fairness measure. As alternatives, the ROBUST researchers have proposed several measures based on the Lorenz curve, which is used in economics, and on the concept of entropy from physics. The paper, which will be presented in June at the Web Science Conference, shows that these measures do not suffer under the drawbacks of the power-law exponent.
This study has used code and datasets from the Koblenz Network Collection (KONECT): http://konect.uni-koblenz.de/
Web Science Conference: http://www.websci12.org/
Fairness on the Web: Alternatives to the Power Law. Jérôme Kunegis, Julia Preusse. In: Proceedings of the Web Science Conference, 2012.