The long-term impact of ranking algorithms in growing networks
share
When we search online for content, we are constantly exposed to rankings. For example, web search results are presented as a ranking, and online bookstores often show us lists of best-selling books. While popularity- and network-based ranking metrics such as degree and Google's PageRank have been extensively studied in previous literature, we still lack a clear understanding of the potential systemic consequences of the adoption of different ranking algorithms. In this work, we fill this gap by introducing a model of network growth where the probability that a node acquires a new connection depends on both its current ranking position and its inherent quality. The model allows us to quantify the ability of a ranking algorithm to detect and promote high-quality content, as well as the heterogeneity of the resulting content popularity distribution. We show that by correcting for the omnipresent age bias of ranking metrics, the resulting networks exhibit a significantly larger agreement between the nodes' quality and their long-term popularity, and larger popularity diversity. Our findings move the first steps toward a model-based understanding of the long-term impact of popularity metrics, and could be used as an informative tool for the design of improved information filtering tools.
READ FULL TEXT