Emergency Risk Communication

A Structural Topic Modelling Analysis of the UK government’s COVID-19 Press Briefings

Abstract

The ongoing coronavirus outbreak has caused a public health emergency of international concern. During public health emergencies, effective risk communication plays an indispensable part in a country’s emergency response. This paper explores the use of Structural Topic Modelling, a machine learning technique that automatically identifies key topics and their content in textual data, in analysing emergency risk communication (ERC) practice at the state level. The data is from the UK government’s COVID-19 press briefings televised between March 2020 and June 2021, totalling approximately 1 million words. The study identifies the prominent topics covered in those briefings as well as their distribution over time, which in turn reflect the UK government’s priorities in handling the public health emergency. Close scrutiny of the use of a selection of key words in context sheds further light on the government’s ERC practice from a linguistic point of view.