Classification of Hashtags’ Growth Based on the Yule–Simon Process

Abstract

A hashtag is a short text for annotation attached to content shared on the Internet. It is known that hashtags increase in their vocabulary and individual cumulative occurrences, and their population dynamics are well approximated by a simple stochastic process, the Yule–Simon process. However, when tags are observed individually on short time scales, there is a macroscopic trend that is different from noise, in which the growth curve may deviate significantly from the approximation. The objective of this study is to quantitatively evaluate the degree of deviation from the Yule–Simon process for each defined time interval. By measuring the degree of this deviation using real data, we confirm that our method is capable of characterizing hashtag growth patterns.