Enhancing Hate Speech Detection on Social Media: A Comparative Analysis of Machine Learning Models and Text Transformation Approaches
#Hate Speech Detection #Machine Learning #BERT #Text Transformation #Social Media #Neural Networks #Content Moderation
📌 Key Takeaways
- Researchers compared traditional and advanced machine learning models for hate speech detection
- BERT models showed superior accuracy in identifying hate speech
- Hybrid models demonstrated improved capabilities in certain scenarios
- New text transformation approaches can convert negative expressions into neutral ones
- The study proposes future directions for more robust hate speech detection systems
📖 Full Retelling
Researchers Saurabh Mishra, Shivani Thakur, and Radhika Mamidi published a comparative analysis of machine learning models for hate speech detection on the arXiv preprint server on February 24, 2026, addressing the growing challenge of harmful content on social media platforms. The study evaluated various machine learning models in identifying hate speech and offensive language, comparing traditional approaches like CNNs and LSTMs with advanced neural network models such as BERT and its derivatives. The researchers also explored hybrid models that combine different architectural features to determine which approaches perform best in detecting problematic content. Their findings revealed that while advanced models like BERT demonstrate superior accuracy due to their deep contextual understanding, hybrid models exhibit improved capabilities in specific scenarios, offering a more nuanced approach to content moderation. Additionally, the paper introduced innovative text transformation techniques that can convert negative expressions into neutral ones, potentially mitigating the impact of harmful content before it reaches users. The authors thoroughly discussed the implications of these findings, highlighting both the strengths and limitations of current technologies and proposing future directions for developing more robust hate speech detection systems that could better protect online communities.
🏷️ Themes
Machine Learning, Hate Speech Detection, Social Media Moderation
📚 Related People & Topics
Social media
Virtual online communities
Social media are new media technologies that facilitate the creation, sharing and aggregation of content (such as ideas, interests, and other forms of expression) amongst virtual communities and networks. Common features include: Online platforms enable users to create and share content and partici...
Machine learning
Study of algorithms that improve automatically through experience
Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of statistical algorithms that can learn from data and generalize to unseen data, and thus perform tasks without explicit instructions. Within a subdiscipline in machine learning, advances i...
Entity Intersection Graph
Connections for Social media:
🌐
Meta
3 shared
👤
Donald Trump
2 shared
👤
Generation Z
2 shared
🌐
Internet safety
2 shared
🌐
TikTok
2 shared
Mentioned Entities
Original Source
--> Computer Science > Computation and Language arXiv:2602.20634 [Submitted on 24 Feb 2026] Title: Enhancing Hate Speech Detection on Social Media: A Comparative Analysis of Machine Learning Models and Text Transformation Approaches Authors: Saurabh Mishra , Shivani Thakur , Radhika Mamidi View a PDF of the paper titled Enhancing Hate Speech Detection on Social Media: A Comparative Analysis of Machine Learning Models and Text Transformation Approaches, by Saurabh Mishra and 2 other authors View PDF Abstract: The proliferation of hate speech on social media platforms has necessitated the development of effective detection and moderation tools. This study evaluates the efficacy of various machine learning models in identifying hate speech and offensive language and investigates the potential of text transformation techniques to neutralize such content. We compare traditional models like CNNs and LSTMs with advanced neural network models such as BERT and its derivatives, alongside exploring hybrid models that combine different architectural features. Our results indicate that while advanced models like BERT show superior accuracy due to their deep contextual understanding, hybrid models exhibit improved capabilities in certain scenarios. Furthermore, we introduce innovative text transformation approaches that convert negative expressions into neutral ones, thereby potentially mitigating the impact of harmful content. The implications of these findings are discussed, highlighting the strengths and limitations of current technologies and proposing future directions for more robust hate speech detection systems. Comments: 32 pages, 24 figures Subjects: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI) ACM classes: I.2.7 Cite as: arXiv:2602.20634 [cs.CL] (or arXiv:2602.20634v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2602.20634 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Saurabh Mishra [...
Read full article at source