Summary
This systematic review examines recent advances (2020 onwards) in automated hate speech detection on digital platforms using artificial intelligence and natural language processing techniques. The authors synthesise existing literature on machine learning and deep learning approaches, review available hate speech datasets, and present empirical implementations comparing classic and deep learning models whilst introducing new metrics for quantifying hatefulness and hatred intensity.
UK applicability
The methodologies and datasets reviewed may inform UK content moderation policy and platform governance, though the paper does not explicitly address UK-specific regulatory contexts (such as Online Safety Bill implementation) or regional linguistic and cultural hate speech patterns.
Key measures
Performance metrics for hate speech detection models (Logistic Regression, LSTM, multi-label architecture); hatefulness quantification; hatred intensity metrics
Outcomes reported
The paper reviews hate speech detection methodologies across textual, multi-modal, and multilingual data modalities, and implements comparative models using Logistic Regression, LSTM, and multi-label architectures. The study derives novel metrics to quantify hatefulness intensity and compares performance across classic machine learning and deep learning techniques.
Topic tags
Dig deeper with Pulse AI.
Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.