Science AI Journal
Computer Science · Research topic

Open research questions in Hate Speech and Cyberbullying Detection

209 gap statements mined from Hate Speech and Cyberbullying Detection papers in our 7.9M-paper local library, which holds 1,313 papers on the topic — drawn mostly from each paper's own stated research gap, future-work, challenge and limitation notes, and its abstract. The ones listed below are a selection still marked open; each names the study that raised it, with a DOI link where the paper has one.

Choosing where to publish on Hate Speech and Cyberbullying Detection? See the ranked Computer Science journals.

What the literature leaves open

  • The lack of a shared practice for detecting obfuscated hate speech in Arabic. The need for a large-scale dataset for Arabic hate speech detection. The limited research on Arabic compared to English.

    ObfXhate: a large-scale Arabic dataset and transformer-based framework for detecting obfuscated hate speech · 2026 · DOI

  • The gap is that current research has not fully explored how administrators at private institutions respond to hate speech incidents. The paper identifies a need for a more nuanced understanding of the relationship between administrator responses and racial inclusion. The study highlights the importance of considering the role of normative understandings about freedom of expression in shaping administrator responses.

    Beyond Legal Constraints: Understanding Barriers to Humanizing Responses in the Aftermath of Hate Speech at Private Universities · 2024 · DOI

  • Platforms that host young users screen messages with commercial text-safety APIs, but independent evidence on how often such systems miss harm or flag harmless text is scarce, especially outside English, and the documentation of the one studied here leaves its false-positive rate unquantified.

    SITO: Sifting a Commercial Text-Safety API for Missed Harm and False Flags in English, Ukrainian and Surzhyk · 2026

  • In this paper, we examine how preference construction can be checked when intermediate decisions are recorded, and which of them remain uncertain.

    Auditing Safety Preference Supervision with Recorded Derivations · 2026

  • Large language models (LLMs) can influence user trust and decision-making through subtle linguistic cues, yet the behavioral effects of manipulative outputs remain poorly quantified.

    When LLMs Manipulate: Measuring Behavioral, Social, and Temporal Effects · 2026

  • While the spread of toxicity on social media has received ample attention, studies offer conflicting expectations and mixed evidence about its contagious nature.

    How social media users adopt the toxic behaviors of ingroup and outgroup accounts · 2025 · 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

  • Future work will explore transformer-based models (BERT, RoBERTa) for context-aware classification, multimodal analysis incorporating image and audio content, graph-based user network analysis for behavioral pattern detection, and federated learning approaches for privacy-preserving model training on distributed social media data.

    DETECTION OF CHILD PREDATORS CYBER HARASSERS ON SOCIAL MEDIA · 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.

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

  • 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 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 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 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

  • Hateful memes require compositional multimodal reasoning to detect harmful intent. Thinking-based MLLMs remain underexplored for hateful meme analysis. The paper must develop a novel approach to improve hateful meme detection.

    Can Thinking Models Think to Detect Hateful Memes? · 2026 · DOI

  • Future research should investigate the long-term effects of online harassment on individuals with mental health challenges. Future research should explore the effectiveness of context-aware interventions and trauma-informed platform designs.

    Adverse Mental Health Events as a Trigger for Online Harassment: Evidence from Online Conversations of Youth Experiencing Psychiatric Hospitalizations · 2026 · DOI

  • The gap in understanding how psychiatric hospitalization may causally impact the likelihood of experiencing online harassment. The gap in understanding the protective role of social support and resilience in reducing the likelihood of harassment.

    Adverse Mental Health Events as a Trigger for Online Harassment: Evidence from Online Conversations of Youth Experiencing Psychiatric Hospitalizations · 2026 · DOI

  • 4, we will present and discuss the four most prevalent open issues (I1-I4) in research as our third contribution in Sec.

    A Survey of Machine Learning Models and Datasets for the Multi-label Classification of Textual Hate Speech in English · 2026 · DOI

  • Traditional moderation systems are inefficient and cannot process the massive volume of data generated every second on social media platforms. Hate speech often appears in implicit, sarcastic, or context-dependent forms, making detection difficult for conventional rule-based systems.

    Design and Implementation of a Comprehensive AI-Driven NLP Framework for Classifying and Preventing Hate Speech in Social Media · 2026 · DOI

  • Our results are based on the Hateful Memes benchmark and may not generalize across domains or languages. Weakly supervised CoT distillation from commercial MLLMs can introduce bias and unfaithful reasoning. GRPO increases computational cost and reliance on closed models may affect reproducibility. Explanation quality is assessed using automatic metrics that only partially reflect human judgment.

    Can Thinking Models Think to Detect Hateful Memes? · 2026 · DOI

  • The study only examined tweets from platform X, - The sample size was limited to 400 tweets, - The study only focused on intentional…

    “She’ll Never Be a Man” A Corpus-Based Forensic Linguistic Analysis of Misgendering Discrimination on X · 2024 · DOI

  • Improved automatic detection systems to better identify and address misgendering in online discourse, - Further empirical investigations to understand and address…

    “She’ll Never Be a Man” A Corpus-Based Forensic Linguistic Analysis of Misgendering Discrimination on X · 2024 · 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, search the Research Gap Finder: the search is free with a free account and lists the papers closest to your topic first. Unlocking that topic (50 credits, charged once) fills the comparison table from our 7.9M-paper local library and writes the gaps from its rows.

Open the Research Gap Finder →

Related topics in Computer Science

Online Learning and Analytics
1,253 open questions
Advanced Neural Network Applications
1,236 open questions
AI in Service Interactions
1,161 open questions
Virtual Reality Applications and Impacts
1,060 open questions
Computational Drug Discovery Methods
1,019 open questions
Quantum Computing Algorithms and Architecture
877 open questions
AI in cancer detection
835 open questions
Topic Modeling
780 open questions
Open questions raised by papers listed on this page include: (1) The lack of a shared practice for detecting obfuscated hate speech in Arabic (“ObfXhate: a large-scale Arabic dataset and transformer-based framework for…”, 2026); (2) 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); (3) Future research should focus on developing more transparent and accountable algorithmic systems (“Social Media Algorithms and Artificial Intelligence: Transformation of the…”, 2026); (4) Traditional approaches have limitations such as poor generalization or lack of interpretability (“A Hybrid Ensemble Framework for Cyberbullying Detection using Multi-Model…”, 2026); (5) Thinking-based MLLMs remain underexplored for hateful meme analysis (“Can Thinking Models Think to Detect Hateful Memes?”, 2026). 209 gap statements have been mined from Hate Speech and Cyberbullying Detection papers in our 7.9M-paper local library, which holds 1,313 papers on the topic; the gaps come from whichever of those papers state one. They are mostly the research gaps the authors state and the papers' abstracts, plus future-work, limitations and challenges passages. The ones listed below are a selection still marked open; each names the study that raised it, with a DOI link where the paper has one, so you can read the original claim in context.

Tools for your next paper

Gaps & Literature SearchFind open questions in any field — 1,100,000+ open gaps mined from our own library.Pre-CheckIs your paper ready? Acceptance odds for top, mid and open/emerging journals, in seconds.Journal FinderA ranked shortlist of target journals from a 17,500-venue index.Duplicate Publication CheckerCheck for prior publication across 7 sources and our local library.Citation GeneratorFreeA DOI or title → APA, MLA, Vancouver, Chicago, BibTeX, RIS. No signup.Graphical Abstract MakerFreeTurn your findings into a colorblind-safe graphical abstract.

How Pre-Check worksHow AI Review works

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

Command palette

Jump anywhere, run any action.