The Structure of Toxic Conversations on Twitter
Open accessThe DOI matches its Crossref record; 7 of the 7 references it cites were found in a registry or an index.
Abstract
Social media platforms promise to enable rich and vibrant conversations online; however, their potential is often hindered by antisocial behaviors. In this paper, we study the relationship between structure and toxicity in conversations on Twitter. We collect 1.18M conversations (58.5M tweets, 4.4M users) prompted by tweets that are posted by or mention major news outlets over one year and candidates who ran in the 2018 US midterm elections over four months. We analyze the conversations at the individual, dyad, and group level. At the individual level, we find that toxicity is spread across many low to moderately toxic users. At the dyad level, we observe that toxic replies are more likely to come from users who do not have any social connection nor share many common friends with the poster. At the group level, we find that toxic conversations tend to have larger, wider, and deeper reply trees, but sparser follow graphs. To test the predictive power of the conversational structure, we consider two prediction tasks. In the first prediction task, we demonstrate that the structural features can be used to predict whether the conversation will become toxic as early as the first ten replies. In the second prediction task, we show that the structural characteristics of the conversation are also predictive of whether the next reply posted by a specific user will be toxic or not. We observe that the structural and linguistic characteristics of the conversations are complementary in both prediction tasks. Our findings inform the design of healthier social media platforms and demonstrate that models based on the structural characteristics of conversations can be used to detect early signs of toxicity and potentially steer conversations in a less toxic direction.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Exposure to ideologically diverse news and opinion on Facebook
Bakshy, E., Messing, S., & Adamic, L. A. (2015) · Science · doi:10.1126/science.aaa1160 · Cited by 3,245
Automated Hate Speech Detection and the Problem of Offensive Language
Davidson, T., Warmsley, D., Macy, M., & Weber, I. (2017) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v11i1.14955 · Cited by 2,610
A Survey on Automatic Detection of Hate Speech in Text
Fortuna, P., & Nunes, S. (2019) · ACM Computing Surveys · doi:10.1145/3232676 · Cited by 1,241 · cited 1×
“These approaches typically analyze the content in isolation and after the fact [21].”
BackgroundThe Risk of Racial Bias in Hate Speech Detection
Sap, M., Card, D., Gabriel, S., Choi, Y., & Smith, N. A. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/p19-1163 · Cited by 817
Ex Machina: Personal Attacks Seen at Scale
Wulczyn, E., Thain, N., & Dixon, L. (2017) · Proceedings of the 26th International Conference on World Wide Web · doi:10.1145/3038912.3052591 · Cited by 636
Fair and Balanced? Quantifying Media Bias through Crowdsourced Content Analysis
Budak, C., Goel, S., & Rao, J. M. (2016) · Public Opinion Quarterly · doi:10.1093/poq/nfw007 · Cited by 336 · cited 1×
“We selected five news outlets that span the political spectrum—NewYork Times and CNN on the left,Wall Street Journal in the middle, and Fox and Breitbart on the right [5, 9]—and have Twitter accounts with a large number of followers.”
Conversations Gone Awry: Detecting Early Signs of Conversational Failure
Zhang, J., Chang, J., Danescu-Niculescu-Mizil, C., Dixon, L., Hua, Y., Taraborelli, D., & Thain, N. (2018) · Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/p18-1125 · Cited by 175 · cited 2×
“Prior work has shown that this approach performs significantly better than bag-of-words and sentiment analysis, and on par with hand-crafted linguistic features designed for this task [53].”
Background
Cited by
Auditing Elon Musk's Impact on Hate Speech and Bots
Hickey, D., Schmitz, M., Fessler, D., Smaldino, P. E., Muric, G., & Burghardt, K. (2023) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v17i1.22222 · Cited by 100
The Platformization of Violence: Toward a Concept of Discursive Toxicity on Social Media
Recuero, R. (2024) · Social Media + Society · doi:10.1177/20563051231224264 · Cited by 47
Evaluating Twitter’s algorithmic amplification of low-credibility content: an observational study
Corsi, G. (2024) · EPJ Data Science · doi:10.1140/epjds/s13688-024-00456-3 · Cited by 45
“Like tweets, Wikipedia comments are largely short and informal, making this model quite suited for the analysis of Twitter data [40, 41].”
Users’ Behavioral and Emotional Response to Toxicity in Twitter Conversations
Aleksandric, A., Saha Roy, S., Pankaj, H., Wilson, G. M., & Nilizadeh, S. (2024) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v18i1.31295 · Cited by 5
Metadata
- Authors
- Saveski, M., Roy, B., & Roy, D.
- Venue
- Proceedings of the Web Conference 2021
- Publisher
- ACM
- Published
- 2021-04-19
- Pages
- 1086-1097
- OpenAlex
- W3155541393
- Cited by
- 83
- References
- 7
- FWCI
- 5.80
- Type
- conference paper
- Language
- en
- Open access
- gold
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: The Structure of Toxic Conversations on Twitter. https://app.cochat.ai/papers/10.1145/3442381.3449861
- The paper
- Saveski, M., Roy, B., & Roy, D. (2021). The Structure of Toxic Conversations on Twitter. Proceedings of the Web Conference 2021. https://doi.org/10.1145/3442381.3449861
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.