The Invasion of Ukraine Viewed through Large-Scale Analysis of TikTok

Abstract

[EN] The vast majority of TikTok analysis done to date has used small, manually-collected datasets. This leaves open the questions of how to do large-scale analysis of phenomena on TikTok and what can be learned about unfolding current events. In this paper, we seek to better understand how such events present themselves on TikTok by conducting a first examination of large-scale user and content dynamics around the invasion of Ukraine in 2022. As this is among the first studies to conduct large-scale (i.e., involving millions of data points) data collection and analysis of TikTok data, our contributions also include insights into best (and less-than-great) practices for such critical research tasks. Furthermore, we have open-sourced our Ukraine-invasion dataset and provide a software library that can be used to collect TikTok data.