EMNLP Findings2021
WIKIBIAS: Detecting Multi-Span Subjective Biases in Language
Yang Zhong, Jingfeng Yang, Wei Xu, Diyi Yang
Abstract
Biases continue to be prevalent in modern text and media, especially subjective bias -a special type of bias that introduces improper attitudes or presents a statement with the presupposition of truth. To tackle the problem of detecting and further mitigating subjective bias, we introduce a manually annotated parallel corpus WIKIBIAS with more than 4,000 sentence pairs from Wikipedia edits. This corpus contains annotations towards both sentencelevel bias types and token-level biased segments. We present systematic analyses of our dataset and results achieved by a set of state-ofthe-art baselines in terms of three tasks: bias classification, tagging biased segments, and neutralizing biased text. We find that current models still struggle with detecting multi-span biases despite their reasonable performances, suggesting that our dataset can serve as a useful research benchmark. We also demonstrate that models trained on our dataset can generalize well to multiple domains such as news and political speeches. 1
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
4 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
Cited by top-tier papers4
- Bias Neutralization in Non-Parallel Texts: A Cyclic Approach with Auxiliary GuidanceKarthic Madanagopal, James CaverleeEMNLP 2023 · 1 citation
- ICLEF: In-Context Learning with Expert Feedback for Explainable Style TransferArkadiy Saakyan, Smaranda MuresanACL 2024
- CoBA: Counterbias Text Augmentation for Mitigating Various Spurious Correlations via Semantic TriplesKyohoon Jin, Juhwan Choi, Jungmin Yun, Junho Lee et al.EMNLP 2025
- CARE: Multilingual Human Preference Learning for Cultural AwarenessGeyang Guo, Tarek Naous, Hiromi Wakaki, Yukiko Nishimura et al.EMNLP 2025
Builds on2
- BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and ComprehensionMike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad et al.ACL 2020 · 1,224 citations
- Politeness Transfer: A Tag and Generate ApproachAman Madaan, Amrith Setlur, Tanmay Parekh, Barnabás Póczos et al.ACL 2020 · 6 citations
Related papers
- SAFARI: Cross-lingual Bias and Factuality Detection in News Media and News ArticlesDilshod Azizov, Zain Muhammad Mujahid, Hilal AlQuabeh, Preslav Nakov et al.EMNLP Findings 2024 · 2 citations
- Cost-Efficient Subjective Task Annotation and Modeling through Few-Shot Annotator AdaptationPreni Golazizian, Alireza Salkhordeh Ziabari, Ali Omrani, Morteza DehghaniEMNLP Findings 2024 · 1 citation
- SWiPE: A Dataset for Document-Level Simplification of Wikipedia PagesPhilippe Laban, Jesse Vig, Wojciech Kryscinski, Shafiq Joty et al.ACL 2023 · 5 citations
- Towards Understanding Gender Bias in Relation ExtractionAndrew Gaut, Tony Sun, Shirlyn Tang, Yuxin Huang et al.ACL 2020 · 7 citations
- Par-ITA: Benchmarking Seq2Seq and LLMs on a Human-Supervised Parallel Corpus for Italian Hyperpartisan NeutralizationMichele Joshua Maggini, Søren Fomsgaard, Michele Maestroni, Gaël Dias et al.ACL 2026
