Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Patterns of Bot Participation and Emotional Influence in Open-Source Development

Matteo Vaccargiu, Riccardo Lai, Maria Ilaria Lunesu, Andrea Pinna, Giuseppe Destefanis

TL;DR

This study investigates how bots participate in Ethereum OSS discussions and whether their interventions shape developer emotions. Using a dataset of 36,875 accounts across ten repositories and a three-stage bot-detection pipeline, the authors identify 105 validated bots and analyze bot versus human participation alongside emotion dynamics detected by a RoBERTa-based 27-emotion classifier. They find bots are neutral on average but influence subsequent human remarks, increasing gratitude, optimism, and admiration while reducing confusion, with changes occurring within about 24 hours. The results reveal that even a small number of bots interact differently with pull requests and issues and can meaningfully affect the tone and timing of developer communication, with implications for bot design and governance in OSS ecosystems.

Abstract

We study how bots contribute to open-source discussions in the Ethereum ecosystem and whether they influence developers' emotional tone. Our dataset covers 36,875 accounts across ten repositories with 105 validated bots (0.28%). Human participation follows a U-shaped pattern, while bots engage in uniform (pull requests) or late-stage (issues) activity. Bots respond faster than humans in pull requests but play slower maintenance roles in issues. Using a model trained on 27 emotion categories, we find bots are more neutral, yet their interventions are followed by reduced neutrality in human comments, with shifts toward gratitude, admiration, and optimism and away from confusion. These findings indicate that even a small number of bots are associated with changes in both timing and emotional dynamics of developer communication.

Patterns of Bot Participation and Emotional Influence in Open-Source Development

TL;DR

This study investigates how bots participate in Ethereum OSS discussions and whether their interventions shape developer emotions. Using a dataset of 36,875 accounts across ten repositories and a three-stage bot-detection pipeline, the authors identify 105 validated bots and analyze bot versus human participation alongside emotion dynamics detected by a RoBERTa-based 27-emotion classifier. They find bots are neutral on average but influence subsequent human remarks, increasing gratitude, optimism, and admiration while reducing confusion, with changes occurring within about 24 hours. The results reveal that even a small number of bots interact differently with pull requests and issues and can meaningfully affect the tone and timing of developer communication, with implications for bot design and governance in OSS ecosystems.

Abstract

We study how bots contribute to open-source discussions in the Ethereum ecosystem and whether they influence developers' emotional tone. Our dataset covers 36,875 accounts across ten repositories with 105 validated bots (0.28%). Human participation follows a U-shaped pattern, while bots engage in uniform (pull requests) or late-stage (issues) activity. Bots respond faster than humans in pull requests but play slower maintenance roles in issues. Using a model trained on 27 emotion categories, we find bots are more neutral, yet their interventions are followed by reduced neutrality in human comments, with shifts toward gratitude, admiration, and optimism and away from confusion. These findings indicate that even a small number of bots are associated with changes in both timing and emotional dynamics of developer communication.
Paper Structure (27 sections, 3 equations, 3 figures)

This paper contains 27 sections, 3 equations, 3 figures.

Table of Contents

  1. Introduction
  2. Related Works
  3. Bot accounts in version control systems
  4. Effectiveness and emotional influence of bot comments in the development workflow
  5. Methodology
  6. Dataset
  7. Project Selection and Ecosystem
  8. Bot Detection
  9. Emotion Detection
  10. Analysis Methods
  11. Temporal Patterns of Bot Participation (RQ1)
  12. Comment Position Within Thread Lifecycle
  13. Response Time Analysis
  14. Emotional Patterns Around Bot Interventions (RQ2)
  15. Baseline Emotional Divergence
  16. ...and 12 more sections

Figures (3)

  • Figure 1: Pull request lifecycle in truffle: human comment density (blue) shows U-shaped pattern; bot markers (orange) show early concentration with continuous activity.
  • Figure 2: Issue lifecycle in truffle: bot activity concentrates late-stage.
  • Figure 3: Neutral emotion probability in bot vs. human comments across all repositories. Bots exhibit significantly higher neutral probability (median 0.76) compared to humans (median 0.47), with minimal overlap between distributions (Mann-Whitney U test: $p < 0.001$, Cliff's Delta $\delta = 0.740$).