Science AI Journal
Computer Science · Research topic

Open research questions in Hate Speech and Cyberbullying Detection

155 unresolved questions extracted from the limitations and future-work sections of 868 Hate Speech and Cyberbullying Detection papers in our library. Each links back to the study that raised it.

What the literature leaves open

  • The study identifies a gap in research on discursive violence against the LGBTQAI+ community on social media. The paper highlights the need to address online violence and promote a safer online environment for the LGBTQAI+ community. The study notes that prior work has not fully explored the impact of de-individuation on online violence against the LGBTQAI+ community.

    Words That Wound: Discursive Violence and the Trolling of LGBTQAI+ Community in Social Media · 2026 · DOI

  • The paper identifies the challenge of algorithmic bias, highlighting that AI systems inevitably reproduce the social, linguistic, and political assumptions present in the data and institutions that produce them. The study notes the challenge of unevenly distributed harms produced by AI systems, disproportionately affecting individuals and communities at the intersection of multiple axes of marginalisation. The research highlights the challenge of developing participatory approaches to AI-based content moderation, which are currently scarce in the field.

    Artificial Intelligence and LGBTQI+phobia: Algorithms of Exclusion or Tools of Inclusion? · 2026 · DOI

  • examining the intersectional dynamics of bias and harm in AI systems, - investigating the effects of AI moderation on mental health, stigma, and well-being, - exploring participatory approaches to AI-based content moderation

    Artificial Intelligence and LGBTQI+phobia: Algorithms of Exclusion or Tools of Inclusion? · 2026 · DOI

  • The paper identifies a gap in the understanding of how the mode of hearing affects the use of adversarial questions in oral hearings. The study notes that prior research has investigated the use of why-fronted questions in oral hearings, but the current study focuses on adversarial questions. The paper highlights the need to understand the use of adversarial questions in oral hearings conducted by the Parole Board.

    Held to account: Comparing adversarial questioning in remote and in-person parole hearings · 2025 · DOI

  • investigating the use of why-fronted questions in oral hearings - researching the effects of remote hearings on the parole process

    Held to account: Comparing adversarial questioning in remote and in-person parole hearings · 2025 · DOI

  • Although research on hate speech is extensive, the systematic characterization of the disseminators who target authoritative news profiles on social platforms remains underexplored.

    Exploring Hate Speech Perpetrators' Profiles on X: An Analysis of User Replies to Spanish Digital News Media Publications · 2026 · DOI

  • While many examined the domestic consequences of hate crimes, little is known about their impact on foreign public perception of the country.

    Anti-Asian Hate Crimes and American Reputation · 2025 · DOI

  • There is a lack of understanding of how humans process misspellings. There is a need for more research on the challenges posed by misspellings in multilingual contexts.

    Misspellings in natural language processing: A survey of recent literature · 2026 · DOI

  • Current work lacks exploration of adversarial misspelling generation and targeted attacks on NLP systems; there is no systematic study of which spelling error patterns are most harmful to different downstream tasks or how to generate minimal misspellings that degrade performance.

    Misspellings in natural language processing: A survey of recent literature · 2026 · DOI

  • The traditional view of online anonymity and its impact on behavior is incomplete. There is a need for a more nuanced understanding of the psychological factors behind toxic online behavior.

    The Psychology Behind Online Anonymity and Toxic Behaviour · 2026 · DOI

  • The paper identifies the paradox of anonymous intimacy (freedom and vulnerability trade-offs including deception risk and diminished real-world social skills) but does not specify measurement instruments, longitudinal study designs, or control group comparisons needed to quantify these outcomes in users of different anonymous platforms or AI-driven companionship systems.

    The Psychology Behind Online Anonymity and Toxic Behaviour · 2026 · DOI

  • Incorporating class-balancing techniques such as SMOTE oversampling or class-weighted loss functions to address class imbalance. Evaluating the system's

    DETECTION OF CHILD PREDATORS CYBER HARASSERS ON SOCIAL MEDIA · 2026 · DOI

  • Traditional detection approaches are inadequate due to the exponential volume of user-generated content and sophisticated evasion tactics. The need for a detection system that can classify user messages as Normal, Harassment, or Predatory with high accuracy.

    DETECTION OF CHILD PREDATORS CYBER HARASSERS ON SOCIAL MEDIA · 2026 · DOI

  • The lack of a central authority in Web 3.0 environments. The 'trustless' environment created by cryptographic anonymity and digital identity systems. The need for a multimodal approach to detect cyberbullying.

    Hybrid Text–Image Fusion Model for Early Detection of Cyberbullying in Online Social Ecosystems · 2026 · DOI

  • While this research provides a robust foundation, the rapidly evolving digital landscape offers several avenues for further exploration: Transition to Vision Transformers (ViT): Future iterations could replace the CNN with Vision Transformers to better capture global dependencies in images, potentially identifying even more subtle bullying cues. Video-Based Detection: Expanding the model (e.g., TikTok/Reels) by incorporating Recurrent Neural Networks (RNNs) or 3D-CNNs to analyze temporal actions. OCR Integration: Implementing Optical Character Recognition (OCR) to extract and analyze text embedded inside images (memes), which is a common loophole for current filters. Decentralized AI Nodes: Exploring the deployment of this model as a lightweight "edge" node in Web 3.0 environments to enable privacy-preserving, localized moderation without central data storage. Multilingual and Cultural Adaptation: Training the NLP stream on diverse dialects and "slang" datasets to ensure the model is effective across different global demographics.

    Hybrid Text–Image Fusion Model for Early Detection of Cyberbullying in Online Social Ecosystems · 2026 · DOI

  • There is a growing need for an automated and intelligent system that can efficiently detect and filter toxic content in real time. Traditional methods for detecting toxic comments rely on manual moderation and simple rule-based systems, which are time-consuming, inconsistent, and unable to handle large volumes of data efficiently.

    YouTube Toxic Comment Classification · 2026 · DOI

  • Algorithmic bias and fairness concerns. The need for transparency and accountability in algorithmic systems. The challenge of scaling content moderation beyond human capability.

    Social Media Algorithms and Artificial Intelligence: Transformation of the Information Space - Positive and Negative Aspects (2023-2024, Facebook, X (Formerly Twitter), TikTok, and YouTube) · 2026 · DOI

  • Future research should focus on developing more transparent and accountable algorithmic systems. There is a need for further study on the impact of algorithmic systems on information diversity and polarization.

    Social Media Algorithms and Artificial Intelligence: Transformation of the Information Space - Positive and Negative Aspects (2023-2024, Facebook, X (Formerly Twitter), TikTok, and YouTube) · 2026 · DOI

  • Cyberbullying poses serious psychological and emotional risks to users. Traditional detection methods are limited in their ability to identify contextual and implicit abusive content. Handling long-range dependencies in text classification tasks.

    A Hybrid Ensemble Framework for Cyberbullying Detection using Multi-Model Consensus and Confidence Weighting · 2026 · DOI

  • Traditional approaches have limitations such as poor generalization or lack of interpretability. Most existing ensemble approaches lack a structured confidence-based aggregation strategy.

    A Hybrid Ensemble Framework for Cyberbullying Detection using Multi-Model Consensus and Confidence Weighting · 2026 · DOI

  • The large number of tweets created per second makes manual content moderation impracticable. The complexity of Twitter language and the need to capture contextual relationships and temporal dependencies. The need to monitor social media platforms in real-time with high efficiency.

    Deep Learning Framework for Prior Identification of Threats in Social Media Interactions · 2026 · DOI

  • The lack of effective methods for detecting and mitigating harmful content in social media interactions. The need for a framework that can proactively detect and mitigate harmful content in social media interactions. The need for a framework that can monitor social media platforms in real-time with high efficiency.

    Deep Learning Framework for Prior Identification of Threats in Social Media Interactions · 2026 · DOI

  • The study identifies the challenge of ambiguous intent or contextual interpretations in gray area cases. The authors highlight the difficulty of developing effective moderation tools and strategies to address the challenges of gray area cases. The paper notes the challenge of balancing the need for efficient moderation with the need for careful consideration of complex cases.

    The Gray Area: Characterizing Moderator Disagreement on Reddit · 2026 · DOI

  • The paper identifies a gap in understanding the complexity of content moderation on Reddit, particularly with regards to disagreements between moderators. The study highlights the need for more effective moderation tools and strategies to address the challenges of gray area cases.

    The Gray Area: Characterizing Moderator Disagreement on Reddit · 2026 · DOI

  • Prior work has not tested whether users' behavioral traces can systematically explain or predict blocking outcomes. The study addresses this gap by analyzing blocking behavior on Bluesky at scale.

    Self-moderation in the Decentralized Era: Decoding Blocking Behavior on Bluesky · 2026 · DOI

Most-cited papers in Hate Speech and Cyberbullying Detection

Most recent work

Find a gap in your own Hate Speech and Cyberbullying Detection sub-topic

This page shows what the Hate Speech and Cyberbullying Detection literature already flags as unresolved. To narrow it to your specific question, run the guided finder — it searches the gap library on demand and checks candidates against 250M+ OpenAlex works.

Open the Research Gap Finder →

Related topics in Computer Science

Online Learning and Analytics
1,106 open questions
AI in Service Interactions
861 open questions
Quantum Computing Algorithms and Architecture
645 open questions
Teaching and Learning Programming
623 open questions
Digital literacy in education
617 open questions
Music Technology and Sound Studies
576 open questions
Virtual Reality Applications and Impacts
496 open questions
Topic Modeling
476 open questions
155 open questions have been extracted from the limitations and future-work passages of 868 Hate Speech and Cyberbullying Detection papers in our library. Each one below links back to the study that raised it, so you can read the original claim in context.

Tools for your next paper

Gaps & Literature SearchFind open questions in any field — 120,000+ open gaps mined from our own library.Pre-CheckIs your paper ready? A Tier 1–5 acceptance probability in seconds.Journal FinderA ranked shortlist of target journals from a 17,500-venue index.Duplicate Publication CheckerCheck prior-publication & salami-slicing risk across 8 sources.Citation GeneratorFreeA DOI or title → APA, MLA, Vancouver, Chicago, BibTeX, RIS. No signup.Graphical Abstract MakerFreeTurn your findings into a colorblind-safe graphical abstract.

Compare the category — Honest roundups of the AI research tools, ours listed alongside the alternatives.

Command palette

Jump anywhere, run any action.