Exploring brand perception in COVID vaccine tweets using TOPSA-NMF model

Abstract

COVID-19 vaccines are primary biomedical preventive measures to combat the pandemic. Social media can have either a positive or negative impact on vaccine uptake, depending on their opinions about vaccines. An algorithm is proposed that can use a linear dimension reduction technique such as non-negative matrix factorization (NMF) to identify topics while using the sentiment lexicon to classify the tweet topics as positive, negative, and neutral. This study aims to determine public attitudes towards each topic of tweets about COVID-19 vaccines.