SOTAVerified

Hate Speech Detection

Hate speech detection is the task of detecting if communication such as text, audio, and so on contains hatred and or encourages violence towards a person or a group of people. This is usually based on prejudice against 'protected characteristics' such as their ethnicity, gender, sexual orientation, religion, age et al. Some example benchmarks are ETHOS and HateXplain. Models can be evaluated with metrics like the F-score or F-measure.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–50 of 507 papers

TitleStatusHype
OPT: Open Pre-trained Transformer Language ModelsCode5
Bryndza at ClimateActivism 2024: Stance, Target and Hate Event Detection via Retrieval-Augmented GPT-4 and LLaMACode4
Personalized Large Language ModelsCode2
ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech DetectionCode2
HateBench: Benchmarking Hate Speech Detectors on LLM-Generated Content and Hate CampaignsCode1
STATE ToxiCN: A Benchmark for Span-level Target-Aware Toxicity Extraction in Chinese Hate Speech DetectionCode1
Multi3Hate: Multimodal, Multilingual, and Multicultural Hate Speech Detection with Vision-Language ModelsCode1
MultiHateClip: A Multilingual Benchmark Dataset for Hateful Video Detection on YouTube and BilibiliCode1
CLIcK: A Benchmark Dataset of Cultural and Linguistic Intelligence in KoreanCode1
TurkishBERTweet: Fast and Reliable Large Language Model for Social Media AnalysisCode1
HARE: Explainable Hate Speech Detection with Step-by-Step ReasoningCode1
K-HATERS: A Hate Speech Detection Corpus in Korean with Target-Specific RatingsCode1
KoMultiText: Large-Scale Korean Text Dataset for Classifying Biased Speech in Real-World Online ServicesCode1
Enhancing social network hate detection using back translation and GPT-3 augmentations during training and test-timeCode1
Enhancing Social Network Hate Detection Using Back Translation and GPT-3 Augmentations During Training and Test-TimeCode1
NLPositionality: Characterizing Design Biases of Datasets and ModelsCode1
Facilitating Fine-grained Detection of Chinese Toxic Language: Hierarchical Taxonomy, Resources, and BenchmarksCode1
HateMM: A Multi-Modal Dataset for Hate Video ClassificationCode1
Interpretable Unified Language CheckingCode1
A Federated Approach for Hate Speech DetectionCode1
Why Is It Hate Speech? Masked Rationale Prediction for Explainable Hate Speech DetectionCode1
Generalizable Implicit Hate Speech Detection Using Contrastive LearningCode1
Public Wisdom Matters! Discourse-Aware Hyperbolic Fourier Co-Attention for Social-Text ClassificationCode1
K-MHaS: A Multi-label Hate Speech Detection Dataset in Korean Online News CommentCode1
A Comprehensive Dataset for German Offensive Language and Conversation AnalysisCode1
Multilingual HateCheck: Functional Tests for Multilingual Hate Speech Detection ModelsCode1
Vietnamese Hate and Offensive Detection using PhoBERT-CNN and Social Media Streaming DataCode1
Entropy-based Attention Regularization Frees Unintended Bias Mitigation from ListsCode1
Large-Scale Hate Speech Detection with Cross-Domain TransferCode1
APEACH: Attacking Pejorative Expressions with Analysis on Crowd-Generated Hate Speech Evaluation DatasetsCode1
Deep Learning for Hate Speech Detection: A Comparative StudyCode1
Few-shot Learning with Multilingual Language ModelsCode1
Zero-shot hashtag segmentation for multilingual sentiment analysisCode1
Hate Speech Detection Based on Sentiment Knowledge SharingCode1
HONEST: Measuring Hurtful Sentence Completion in Language ModelsCode1
AraCOVID19-MFH: Arabic COVID-19 Multi-label Fake News and Hate Speech Detection DatasetCode1
Detecting Hate Speech with GPT-3Code1
A Large-scale Dataset for Hate Speech Detection on Vietnamese Social Media TextsCode1
Learning from the Worst: Dynamically Generated Datasets to Improve Online Hate DetectionCode1
HateCheck: Functional Tests for Hate Speech Detection ModelsCode1
Detecting Hate Speech in Multi-modal MemesCode1
HateXplain: A Benchmark Dataset for Explainable Hate Speech DetectionCode1
Toxic Language Detection in Social Media for Brazilian Portuguese: New Dataset and Multilingual AnalysisCode1
ETHOS: an Online Hate Speech Detection DatasetCode1
BEEP! Korean Corpus of Online News Comments for Toxic Speech DetectionCode1
Aggression and Misogyny Detection using BERT: A Multi-Task ApproachCode1
Deep Learning Models for Multilingual Hate Speech DetectionCode1
Classification Benchmarks for Under-resourced Bengali Language based on Multichannel Convolutional-LSTM NetworkCode1
Multilingual Twitter Corpus and Baselines for Evaluating Demographic Bias in Hate Speech RecognitionCode1
DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighterCode1
Show:102550
← PrevPage 1 of 11Next →
All datasetsEthos BinaryHateXplainEthos MultiLabelWaseem et al., 2018AbusEvalAutomatic Misogynistic IdentificationHateMMHatEvalOffensEval 2019ToLD-Brbajer_danish_misogynyDKhate

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1BiLSTM + static BEF1-score0.8—Unverified
2BERTF1-score0.79—Unverified
3BiLSTM+Attention+FTF1-score0.77—Unverified
4OPT-175B (few-shot)F1-score0.76—Unverified
5CNN+Attention+FT+GVF1-score0.74—Unverified
6OPT-175B (one-shot)F1-score0.71—Unverified
7OPT-175B (zero-shot)F1-score0.67—Unverified
8SVMF1-score0.66—Unverified
9Random ForestsF1-score0.64—Unverified
10Davinci (zero-shot)F1-score0.63—Unverified
#ModelMetricClaimedVerifiedStatus
1BERT-MRPAUROC0.86—Unverified
2BERT-RPAUROC0.85—Unverified
3BERT-HateXplain [Attn]AUROC0.85—Unverified
4BERT-HateXplain [LIME]AUROC0.85—Unverified
5BERT [Attn]AUROC0.84—Unverified
6BiRNN-HateXplain [Attn]AUROC0.81—Unverified
7BiRNN-Attn [Attn]AUROC0.8—Unverified
8CNN-GRU [LIME]AUROC0.79—Unverified
9BiRNN [LIME]AUROC0.77—Unverified
10XG-HSI-BERTAccuracy0.75—Unverified
#ModelMetricClaimedVerifiedStatus
1MLARAMHamming Loss0.29—Unverified
2MLkNNHamming Loss0.16—Unverified
3Binary RelevanceHamming Loss0.14—Unverified
4Neural Classifier ChainsHamming Loss0.13—Unverified
5Neural Binary RelevanceHamming Loss0.11—Unverified
#ModelMetricClaimedVerifiedStatus
1Mozafari et al., 2019AAA50.94—Unverified
2SVMAAA46.51—Unverified
3Kennedy et al., 2020AAA45.5—Unverified
#ModelMetricClaimedVerifiedStatus
1HateBERTMacro F10.74—Unverified
2BERTMacro F10.72—Unverified
#ModelMetricClaimedVerifiedStatus
1mBertAccuracy0.83—Unverified
2Logistic RegressionAccuracy0.7—Unverified
#ModelMetricClaimedVerifiedStatus
1HXP + CLAP + CLIPTEST F1 (macro)0.85—Unverified
2BERT + ViT + MFCCTEST F1 (macro)0.79—Unverified
#ModelMetricClaimedVerifiedStatus
1HateBERTMacro F10.49—Unverified
2BERTMacro F10.48—Unverified
#ModelMetricClaimedVerifiedStatus
1HateBERTMacro F10.81—Unverified
2BERTMacro F10.8—Unverified
#ModelMetricClaimedVerifiedStatus
1Multilingual BERTF1-score0.75—Unverified
2AutoMLF1-score0.74—Unverified
#ModelMetricClaimedVerifiedStatus
1AOM mBERTF10.85—Unverified
#ModelMetricClaimedVerifiedStatus
1BaselineF10.7—Unverified
#ModelMetricClaimedVerifiedStatus
1RoBERTa-large-STMacro F180.7—Unverified
#ModelMetricClaimedVerifiedStatus
1Baseline BERT (task A)F10.77—Unverified