Papers › AAA: Fair Evaluation for Abuse Detection Systems Wanted
AAA: Fair Evaluation for Abuse Detection Systems Wanted
Agostina Calabrese, Michele Bevilacqua, Björn Ross, Rocco Tripodi, Roberto Navigli
User-generated web content is rife with abusive language that can harm others and discourage participation. Thus, a primary research aim is to develop abuse detection systems that can be used to alert and support human moderators of online communities. Such systems are notoriously hard to develop and evaluate. Even when they appear to achieve satisfactory performance on current evaluation metrics, they may fail in practice on new data. This is partly because datasets commonly used in this field suffer from selection bias, and consequently, existing supervised models overrely on cue words such as group identifiers (e.g., gay and black) which are not inherently abusive. Although there are attempts to mitigate this bias, current evaluation metrics do not adequately quantify their progress. In this work, we introduce Adversarial Attacks against Abuse (AAA), a new evaluation strategy and associated metric that better captures a model’s performance on certain classes of hard-to-classify microposts, and for example penalises systems which are biased on low-level lexical features. It does so by adversarially modifying the model developer’s training and test data to generate plausible test samples dynamically. We make AAA available as an easy-to-use tool, and show its effectiveness in error analysis by comparing the AAA performance of several state-of-the-art models on multiple datasets. This work will inform the development of detection systems and contribute to the fight against abusive language online.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Hate Speech Detection | Waseem et al., 2018 | Mozafari et al., 2019 | AAA | 50.94 | #1 of 3 | Archive leaderboard | report |
| Hate Speech Detection | Waseem et al., 2018 | Mozafari et al., 2019 | F1 (micro) | 84.42 | #1 of 3 | Archive leaderboard | report |
| Hate Speech Detection | Waseem et al., 2018 | SVM | AAA | 46.51 | #2 of 3 | Archive leaderboard | report |
| Hate Speech Detection | Waseem et al., 2018 | SVM | F1 (micro) | 82.18 | #2 of 3 | Archive leaderboard | report |
| Hate Speech Detection | Waseem et al., 2018 | Kennedy et al., 2020 | AAA | 45.50 | #3 of 3 | Archive leaderboard | report |
| Hate Speech Detection | Waseem et al., 2018 | Kennedy et al., 2020 | F1 (micro) | 82.18 | #3 of 3 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections