Analyzing Sentiments and Topics on Twitter Towards Rising Cost of Living

EasyChair Preprint no. 10605

Abstract

In September 2022, the United Kingdom experienced an un- precedented 40-year high in its inflation rate, resulting in a cost of living crisis that has significantly impacted British citizens. To assess public opinion on this issue, we developed a social media analytics pipeline to collect and analyze microblogs posted on Twitter. Our primary objective was to conduct sentiment analysis on the collected tweets to determine the dominant sentiment towards the topic of the cost of living. Addi- tionally, we performed named entity recognition to identify the entities most frequently mentioned and used topic modeling to uncover the most discussed topics. Our approach employed a hybrid sentiment analysis method that utilized three lexicons for preliminary tweet labeling and fine-tuned a RoBERTa model. Our results demonstrate the superior ef- fectiveness of our methods, which provided an in-depth analysis of the cost-of-living situation in the UK.

Keyphrases: Rising Cost of Living, Sentiment Analysis, text mining, Topic Modelling

@Booklet{EasyChair:10605,
  author = {Yanyi Li and Nian Ran and Yifu Chen and Renhua Zhou and Riza Batista-Navarro},
  title = {Analyzing Sentiments and Topics on Twitter Towards Rising Cost of Living},
  howpublished = {EasyChair Preprint no. 10605},

  year = {EasyChair, 2023}}