Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

OneLove beyond the field -- A few-shot pipeline for topic and sentiment analysis during the FIFA World Cup in Qatar

Christoph Rauchegger, Sonja Mei Wang, Pieter Delobelle

TL;DR

This paper tackles understanding public discourse around the OneLove armband during the 2022 FIFA World Cup in Qatar by combining topic modeling and few-shot sentiment analysis on German Tweets. It leverages BERTopic for topic discovery and in-context learning with large language models (e.g., Mistral-7B, GPT-3.5/4) to infer stance, comparing zero-shot and few-shot setups and validating against a human-annotated gold standard. The authors demonstrate that a few-shot pipeline can quickly adapt to unfolding events, with prompting and translation strategies boosting performance and Mistral achieving competitive accuracy while remaining cost-efficient relative to GPT-4. They also show a temporal shift from armband-specific discourse to broader sports politics and compare social media signals to surveys, highlighting the value and limitations of each source for crisis monitoring and public sentiment analysis.

Abstract

The FIFA World Cup in Qatar was discussed extensively in the news and on social media. Due to news reports with allegations of human rights violations, there were calls to boycott it. Wearing a OneLove armband was part of a planned protest activity. Controversy around the armband arose when FIFA threatened to sanction captains who wear it. To understand what topics Twitter users Tweeted about and what the opinion of German Twitter users was towards the OneLove armband, we performed an analysis of German Tweets published during the World Cup using in-context learning with LLMs. We validated the labels on human annotations. We found that Twitter users initially discussed the armband's impact, LGBT rights, and politics; after the ban, the conversation shifted towards politics in sports in general, accompanied by a subtle shift in sentiment towards neutrality. Our evaluation serves as a framework for future research to explore the impact of sports activism and evolving public sentiment. This is especially useful in settings where labeling datasets for specific opinions is unfeasible, such as when events are unfolding.

OneLove beyond the field -- A few-shot pipeline for topic and sentiment analysis during the FIFA World Cup in Qatar

TL;DR

This paper tackles understanding public discourse around the OneLove armband during the 2022 FIFA World Cup in Qatar by combining topic modeling and few-shot sentiment analysis on German Tweets. It leverages BERTopic for topic discovery and in-context learning with large language models (e.g., Mistral-7B, GPT-3.5/4) to infer stance, comparing zero-shot and few-shot setups and validating against a human-annotated gold standard. The authors demonstrate that a few-shot pipeline can quickly adapt to unfolding events, with prompting and translation strategies boosting performance and Mistral achieving competitive accuracy while remaining cost-efficient relative to GPT-4. They also show a temporal shift from armband-specific discourse to broader sports politics and compare social media signals to surveys, highlighting the value and limitations of each source for crisis monitoring and public sentiment analysis.

Abstract

The FIFA World Cup in Qatar was discussed extensively in the news and on social media. Due to news reports with allegations of human rights violations, there were calls to boycott it. Wearing a OneLove armband was part of a planned protest activity. Controversy around the armband arose when FIFA threatened to sanction captains who wear it. To understand what topics Twitter users Tweeted about and what the opinion of German Twitter users was towards the OneLove armband, we performed an analysis of German Tweets published during the World Cup using in-context learning with LLMs. We validated the labels on human annotations. We found that Twitter users initially discussed the armband's impact, LGBT rights, and politics; after the ban, the conversation shifted towards politics in sports in general, accompanied by a subtle shift in sentiment towards neutrality. Our evaluation serves as a framework for future research to explore the impact of sports activism and evolving public sentiment. This is especially useful in settings where labeling datasets for specific opinions is unfeasible, such as when events are unfolding.
Paper Structure (12 sections, 3 figures, 1 table)

This paper contains 12 sections, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. Method
  4. Topic modeling
  5. Few-shot sentiment analysis
  6. Discussion
  7. Opinions on the OneLove armband.
  8. Few-shot analysis pipeline.
  9. Comparison to Surveys
  10. Conclusion
  11. Stopwords
  12. Prompts

Figures (3)

  • Figure 1: Topic modeling and sentiment analysis pipeline.
  • Figure 2: Timeline of discussed topics related to OneLove. The topics are found by topic modeling (\ref{['ss:topic-modeling']}) and linked to OneLove with manual annotation of a subset of Tweets.
  • Figure 3: Sentiment of Tweets related to OneLove.