Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Sticker is Worth a Thousand Words: Characterizing the Use of Stickers in WhatsApp Political Groups in Brazil

Philipe Melo, João M. M. Couto, Daniel Kansaon, Vitor Mafra, Júlio C. S. Reis, Fabrício Benevenuto

TL;DR

This work provides a large-scale, multi-method examination of sticker use in Brazilian WhatsApp political groups during the 2022 election. It combines perceptual hashing and DBSCAN-based visual clustering with a group-graph approach to reveal how stickers reflect partisan dynamics and social context. The analysis uncovers strong political alignment signals, two dominant polarized communities, and substantial abusive and NSFW content that raises safety concerns. The findings underscore the need for better moderation and suggest extending such analyses to other platforms to understand the broader implications of sticker-based political communication.

Abstract

With the increasing use of smartphones, instant messaging platforms turned into important communication tools. According to WhatsApp, more than 100 billion messages are sent each day on the app. Communication on these platforms has allowed individuals to express themselves in other types of media, rather than simple text, including audio, videos, images, and stickers. Particularly, stickers are a new multimedia format that emerged with messaging apps, promoting new forms of interactions among users, especially in the Brazilian context, transcending their role as a mere form of humor to become a key element in political strategy. In this regard, we investigate how stickers are being used, unveiling unique characteristics that these media bring to WhatsApp chats and the political use of this new media format. To achieve that, we collected a large sample of messages from WhatsApp public political discussion groups in Brazil and analyzed the sticker messages shared in this context

A Sticker is Worth a Thousand Words: Characterizing the Use of Stickers in WhatsApp Political Groups in Brazil

TL;DR

This work provides a large-scale, multi-method examination of sticker use in Brazilian WhatsApp political groups during the 2022 election. It combines perceptual hashing and DBSCAN-based visual clustering with a group-graph approach to reveal how stickers reflect partisan dynamics and social context. The analysis uncovers strong political alignment signals, two dominant polarized communities, and substantial abusive and NSFW content that raises safety concerns. The findings underscore the need for better moderation and suggest extending such analyses to other platforms to understand the broader implications of sticker-based political communication.

Abstract

With the increasing use of smartphones, instant messaging platforms turned into important communication tools. According to WhatsApp, more than 100 billion messages are sent each day on the app. Communication on these platforms has allowed individuals to express themselves in other types of media, rather than simple text, including audio, videos, images, and stickers. Particularly, stickers are a new multimedia format that emerged with messaging apps, promoting new forms of interactions among users, especially in the Brazilian context, transcending their role as a mere form of humor to become a key element in political strategy. In this regard, we investigate how stickers are being used, unveiling unique characteristics that these media bring to WhatsApp chats and the political use of this new media format. To achieve that, we collected a large sample of messages from WhatsApp public political discussion groups in Brazil and analyzed the sticker messages shared in this context
Paper Structure (11 sections, 10 figures, 3 tables)

This paper contains 11 sections, 10 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. Methodology
  4. Data Collection
  5. Processing Sticker Data
  6. Stickers as a distinctive form of media
  7. Stickers Characterization
  8. Clustering by Visual Similarity
  9. Political Alignment of Stickers
  10. Abusive Stickers on WhatsApp
  11. Conclusion

Figures (10)

  • Figure 1: Cumulative Distribution Function (CDF) of stickers sent per groups and users compared to image and text.
  • Figure 2: Cumulative Distribution Function (CDF) of total shares and forwarding per sticker and image medias.
  • Figure 3: UMAP visualization of all stickers from WhatsApp dataset.
  • Figure 4: Example of a cluster of stickers representing emojis.
  • Figure 5: Example of the cluster with meme sticker template with small variations.
  • ...and 5 more figures