Identify Hate Speech Research Articles

Understanding hate speech: the HateInsights dataset and model interpretability

The persistence of hate speech continues to pose an obstacle in the realm of online social media. Despite the continuous evolution of advanced models for identifying hate speech, the critical dimensions of interpretability and explainability have not received proportional scholarly attention. In this article, we introduce the HateInsights dataset, a groundbreaking benchmark in the field of hate speech datasets, encompassing diverse aspects of this widespread issue. Within our dataset, each individual post undergoes thorough annotation from dual perspectives: firstly, conforming to the established 3-class classification paradigm that includes hate speech, offensive language, and normal discourse; secondly, incorporating rationales that outline specific segments of a post supporting the assigned label (categorized as hate speech, offensive language, or normal discourse). Our exploration yields a significant finding by harnessing cutting-edge state-of-the-art models: even models demonstrating exceptional proficiency in classification tasks yield suboptimal outcomes in crucial explainability metrics, such as model plausibility and faithfulness. Furthermore, our analysis underscores a promising revelation concerning models trained using human-annotated rationales. To facilitate scholarly progress in this realm, we have made both our dataset and codebase accessible to fellow researchers. This initiative aims to encourage collaborative involvement and inspire the advancement of the hate speech detection approach characterized by increased transparency, clarity, and fairness.

Utilizing Support Vector Machines to Detect Hate Speech on Social Media

This paper presents a novel application of Support Vector Machines (SVM) in developing automated systems to detect hate speech on social media platforms, addressing a critical need for scalable solutions to enhance online safety and societal cohesion. By leveraging SVM's proven efficacy in managing high-dimensional data and optimizing the balance between precision and recall, the study offers a comprehensive methodology that includes data collection, preprocessing, model training, deployment, and evaluation. The results demonstrate robust average performance across key metrics, affirming the model's reliability in accurately identifying hate speech while minimizing false positives. This research advances the field by showcasing the practical and theoretical contributions of SVM in automated hate speech detection, highlighting its potential to significantly improve content moderation practices. The findings underscore the necessity for ongoing refinement of detection systems and collaborative efforts among researchers, technology firms, and policymakers to create more inclusive online environments that promote respectful discourse and community well-being.

A Hybrid Deep BiLSTM-CNN for Hate Speech Detection in Multi-social media

Nowadays, means of communication among people have changed due to advancements in information technology and the rise of online multi-social media. Many people express their feelings, ideas, and emotions on social media sites such as Instagram, Twitter, Gab, Reddit, Facebook, and YouTube. However, people have misused social media to send hateful messages to specific individuals or groups to create chaos. For various governance authorities, manually identifying hate speech on various social media platforms is a difficult task to avoid such chaos. In this study, a hybrid deep-learning model, where bidirectional long short-term memory (BiLSTM) and convolutional neural network (CNN) are used to classify hate speech in textual data, is proposed. This model incorporates a GLOVE-based word embedding approach, dropout, L2 regularization, and global max pooling to get impressive results. Further, the proposed BiLSTM-CNN model has been evaluated on various datasets to achieve state-of-the-art performance that is superior to the traditional and existing machine learning methods in terms of accuracy, precision, recall, and F1-score.

Development of Pidgin English Hate Speech Classification System for Social Media

With the widespread use of social media, people from all walks of life—individuals, friends, family, public and private organizations, business communities, states, and entire nations—are exchanging information in various formats, including text, messages, audio, video, cartons, and pictures. Social media also facilitates the distribution and propagation of hate speech, despite the immense benefits of knowledge sharing through these platforms. The purpose of this work was to construct a text-based, Pidgin English hate speech classification system (HSCS) in social media, taking into account the alarming rate at which hate speech is shared and propagated on social media, as well as the negative effects of hate speech on society. We used text data sets in Pidgin English that were taken from Twitter and Facebook (3,153). To train the Support Vector Machine (SVM) text classifier to identify hate speech in Pidgin English, 70% of the Pidgin English data set was annotated. The SVM classifier's performance was tested and assessed using the remaining thirty percent of the Pidgin English text data set. The test set findings' confusion matrix, as determined by the HSCS performance evaluation, was 62.04%, 64.42%, 0.7541, 0.6947, and 0.64 in terms of accuracy, precision, recall, F1-score, and Receiver Operating Characteristics (ROC) curve. When HSCS was compared to other Machine Learning (ML) classifiers, such as Logistic Regression (LR), Random Forest (RF), and Naive Bayes, the results showed that LR had accuracy and precision of 61.51% and 63.89%, RF had 54.88% and 50.65%, and Naive Bayes had 61.51% and 63.89%.

Twitter Sentiment Analysis Using Machine Learning

In an age of social media, online forums, and chats, cyberbullying is a prevalent issue. On Twitter (now X), approximately 500 million tweets are shared per day (Antonakaki et.al., 2021). It is the job of the moderators to ensure these tweets follow standard community guidelines. However, the sheer number of tweets makes it difficult to sort manually and ensure they are following protocol. Sentiment analysis and machine learning algorithms can be used to classify these texts automatically as positive or negative. Normally, these machine learning models are much more efficient and may provide higher accuracy rates in identifying hate speech in Twitter. In this paper, we are exploring the use of five classical machine learning algorithms to classify Twitter hate speech as neutral, racist, or sexist. Model performance was compared after using raw tweet data versus pre-processed tweets through data cleanup. Furthermore, we highlight two methods to deal with imbalanced datasets to improve the prediction rates. Overall, we were able to achieve a 96% accuracy in correctly classifying tweets into the different labels.

Automatic Identification of Hate Speech - A Case-Study of alt-Right YouTube Videos.

Identifying hate speech (HS) is a central concern within online contexts. Current methods are insufficient for efficient preemptive HS identification. In this study, we present the results of an analysis of automatic HS identification applied to popular alt-right YouTube videos. This essay describes methodological challenges of automatic HS detection. The case study concerns data on a formative segment of contemporary radical right discourse. Our purpose is twofold. (1) To outline an interdisciplinary mixed-methods approach for using automated identification of HS. This bridges the gap between technical research on the one hand (such as machine learning, deep learning, and natural language processing, NLP) and traditional empirical research on the other. Regarding alt-right discourse and HS, we ask: (2) What are the challenges in identifying HS in popular alt-right YouTube videos? The results indicate that effective and consistent identification of HS communication necessitates qualitative interventions to avoid arbitrary or misleading applications. Binary approaches of hate/non-hate speech tend to force the rationale for designating content as HS. A context-sensitive qualitative approach can remedy this by bringing into focus the indirect character of these communications. The results should interest researchers within social sciences and the humanities adopting automatic sentiment analysis and for those analysing HS and radical right discourse. Automatic identification or moderation of HS cannot account for an evolving context of indirect signification. This study exemplifies a process whereby automatic hate speech identification could be utilised effectively. Several methodological steps are needed for a useful outcome, with both technical quantitative processing and qualitative analysis being vital to achieve meaningful results. With regard to the alt-right YouTube material, the main challenge is indirect framing. Identification demands orientation in the broader discursive context and the adaptation towards indirect expressions renders moderation and suppression ethically and legally precarious.

Repairing the harm: Toward an algorithmic reparations approach to hate speech content moderation

Content moderation algorithms influence how users understand and engage with social media platforms. However, when identifying hate speech, these automated systems often contain biases that can silence or further harm marginalized users. Recently, scholars have offered both restorative and transformative justice frameworks as alternative approaches to platform governance to mitigate harms caused to marginalized users. As a complement to these recent calls, in this essay, I take up the concept of reparation as one substantive approach social media platforms can use alongside and within these justice frameworks to take actionable steps toward addressing, undoing and proactively preventing the harm caused by algorithmic content moderation. Specifically, I draw on established legal and legislative reparations frameworks to suggest how social media platforms can reconceptualize algorithmic content moderation in ways that decrease harm to marginalized users when identifying hate speech. I argue that the concept of reparations can reorient how researchers and corporate social media platforms approach content moderation, away from capitalist impulses and efficiency and toward a framework that prioritizes creating an environment where individuals from marginalized communities feel safe, protected and empowered.

THAR- Targeted Hate Speech Against Religion: A high-quality Hindi-English code-mixed Dataset with the Application of Deep Learning Models for Automatic Detection

During the last decade, social media has gained significant popularity as a medium for individuals to express their views on various topics. However, some individuals also exploit the social media platforms to spread hatred through their comments and posts, some of which target individuals, communities or religions. Given the deep emotional connections people have to their religious beliefs, this form of hate speech can be divisive and harmful, and may result in issues of mental health as social disorder. Therefore, there is a need of algorithmic approaches for the automatic detection of instances of hate speech. Most of the existing studies in this area focus on social media content in English, and as a result several low-resource languages lack computational resources for the task. This study attempts to address this research gap by providing a high-quality annotated dataset designed specifically for identifying hate speech against religions in the Hindi-English code-mixed language. This dataset “Targeted Hate Speech Against Religion” (THAR)) consists of 11,549 comments and has been annotated by five independent annotators. It comprises two subtasks: (i) Subtask-1 (Binary classification), (ii) Subtask-2 (multi-class classification). To ensure the quality of annotation, the Fleiss Kappa measure has been employed. The suitability of the dataset is then further explored by applying different standard deep learning, and transformer-based models. The transformer-based model, namely Multilingual Representations for Indian Languages (MuRIL), is found to outperform the other implemented models in both subtasks, achieving macro average and weighted average F1 scores of 0.78 and 0.78 for Subtask-1, and 0.65 and 0.72 for Subtask-2, respectively. The experimental results obtained not only confirm the suitability of the dataset but also advance the research towards automatic detection of hate speech, particularly in the low-resource Hindi-English code-mixed language.

Hate Speech and Abusive Language and Abusive Language Detection in Twitter using Machine Learning

Twitter's central goal is to enable everybody to make and share thoughts and data, and to communicate their suppositions and convictions without boundaries. Twitter's job is to serve the public discussion, which requires portrayal of a different scope of points of view. Yet, it does not advance viciousness against or straightforwardly assault or undermine others based on race, nationality, public cause, rank, sexual direction, age, inability, or genuine illness. Hate speech and abusive language can hurt a person or a community. So, it is not appropriate to use hate speech and abusive language. Now, due to increase in social media usage, hate speech and abusive language is very commonly used on these platforms. So, it is not possible to identify hate speech and abusive languagees manually. So, it is essential to develop an automated hate speech and abusive language detection model and this research work shows different approaches of Natural Language Processing for classification of Hate speech and abusive language through Machine Learning Algorithms.

Comparative Analysis of NLP Techniques for Hate Speech Classification in Online Communications

This research aimed to compare the effectiveness of two Natural Language Processing (NLP) techniques—SpaCy's word embeddings and Sklearn's TF-IDF vectorization—in identifying hate speech within online comments. Utilizing a balanced dataset, each model was meticulously assessed on its ability to classify comments as 'hateful' or 'non-hateful'. The evaluation metrics employed were precision, recall, F1-score, and overall accuracy. The model using SpaCy's word embeddings achieved an accuracy of 65%, with equal precision and recall for both classes. The Sklearn's TF-IDF vectorization model, however, demonstrated superior performance with an overall accuracy of 75% and an enhanced ability to correctly identify hateful comments, evidenced by a 77% recall rate. This suggests that the TF-IDF model is more adept at discerning nuanced expressions of hate speech. The study's findings highlight the critical role of vectorization methods in the field of automated content moderation and stress the importance of continued innovation and model adaptation to effectively manage the evolving nature of online hate speech.

CROA-based feature selection with BERT model for detecting the offensive speech in Twitter data

<p>Online hate speech has flourished on social networking sites due to the widespread availability of mobile computers and other Web knowledge. Extensive research has shown that online exposure to hate speech has real-world effects on marginalized communities. Research into methods of automatically identifying hate speech has garnered significant attention. Hate speech can affect any demographic, while some populations are more vulnerable than others. Relying solely on progressive learning is insufficient for achieving the goal of automatic hate speech identification. It need access to large amounts of labelled data to train a model. Inaccurate statistics on hate speech and preconceived notions have been the biggest obstacles in the field of hate speech research for a long time. This research provides a novel strategy for meeting these needs by combining a transfer-learning attitude-based BERT (Bidirectional Encoder Representations from Transformers) with a coral reef optimization-based approach (CROA). A feature selection (FC) optimization strategy for coral reefs, a coral reefs optimization method mimics coral behaviours for reef location and development. We might think of each potential answer to the problem as a coral trying to establish itself in the reefs. The results are refined at each stage by applying specialized operators from the coral reefs optimization algorithm. When everything is said and done, the optimal solution is chosen. We also use a cutting-edge fine-tuning method based on transfer learning to assess BERT’s ability to recognize hostile contexts in social media communications. The paper evaluates the proposed approach using Twitter datasets tagged for racist, sexist, homophobic, or otherwise offensive content. The numbers show that our strategy achieves 5%–10% higher precision and recall compared to other approaches.</p>

Identifying Hate Speech Through Syntax Dependency Graph Convolution and Sentiment Knowledge Transfer

Identifying Hate Speech Through Syntax Dependency Graph Convolution and Sentiment Knowledge Transfer

Battling cultural bias within hate speech detection: An experimental correlation analysis

Hate speech detection systems have become essential in the advancing digital world. They limit the dissemination of hateful and offensive language online. However, the machine learning algorithms that provide the basis for these systems struggle to identify hate speech versus clean speech within a cultural context, allowing the growth of cultural bias. Though previous methods had aimed to mitigate the cultural bias of a machine learning model, we attempted to find a new understanding with regard to cultural bias. This study sought to determine a correlation between increasing the amount of cultural speech used to train the machine learning model and the model’s cultural bias when classifying hate speech and clean speech. Additionally, we hypothesized that increasing the cultural weight of a training dataset would mitigate the cultural bias. To test this hypothesis, we created a unique method named Categorial Weighted Training (CaWT), derived from multiple other methods of previous researchers, to identify a correlation. CaWT involved the creation of multiple culturally-weighted training datasets and training a machine learning algorithm against them. From this, the results illustrated minimal correlation between the cultural weight of a training dataset and the model’s performance on cultural speech. However, a significant negative correlation exists between the cultural weight and the model’s performance on non-cultural speech. This implies that increasing the cultural weight does not affect the model’s cultural bias but decreases the model’s performance on non-cultural speech, suggesting that a lower cultural weight is ideal within the limitations of our research.

Autoencoder-Based Feature Extraction for Identifying Hate Speech Spreaders in Social Media

Hate speech on social media has become a big problem, making regular users very upset and giving victims depression and suicidal thoughts. Early identification of the user spreading this type of hate speech may be a better solution, allowing hate speech to be stopped at source. In this article, we attempt to identify these hate speech spreaders by finding a representation for each user. Each user’s comments are aggregated and fed to an auto-encoder to train it. The encoder part of the auto-encoder is used to get an encoded vector for each user. The encoded vector is used with different machine learning (ML) classifiers to determine if a user is spreading hate speech. The proposed model was tested using the dataset released by PAN 2021 (https://pan.webis.de/data.html) hate speech spreader profiling competition in English and Spanish. The experimental results show that support vector machine (SVM) with encoded vectors as features outperforms existing models with an accuracy of 92% for both English and Spanish dataset. The proposed features extraction technique is found to be equally effective at identifying fake news spreaders on fake news datasets provided by PAN 2020 yielding accuracy values of 95% and 83% for English and Spanish, respectively.

Detection of cyberhate speech towards female sport in the Arabic Xsphere.

The recent rapid growth in the number of Saudi female athletes and sports enthusiasts' presence on social media has exposed them to gender-hate speech and discrimination. Hate speech, a harmful worldwide phenomenon, can have severe consequences. Its prevalence in sports has surged alongside the growing influence of social media, with X serving as a prominent platform for the expression of hate speech and discriminatory comments, often targeting women in sports. This research combines two studies that explores online hate speech and gender biases in the context of sports, proposing an automated solution for detecting hate speech targeting women in sports on platforms like X, with a particular focus on Arabic, a challenging domain with limited prior research. In Study 1, semi-structured interviews with 33 Saudi female athletes and sports fans revealed common forms of hate speech, including gender-based derogatory comments, misogyny, and appearance-related discrimination. Building upon the foundations laid by Study 1, Study 2 addresses the pressing need for effective interventions to combat hate speech against women in sports on social media by evaluating machine learning (ML) models for identifying hate speech targeting women in sports in Arabic. A dataset of 7,487 Arabic tweets was collected, annotated, and pre-processed. Term frequency-inverse document frequency (TF-IDF) and part-of-speech (POS) feature extraction techniques were used, and various ML algorithms were trained Random Forest consistently outperformed, achieving accuracy (85% and 84% using TF-IDF and POS, respectively) compared to other methods, demonstrating the effectiveness of both feature sets in identifying Arabic hate speech. The research contribution advances the understanding of online hate targeting Arabic women in sports by identifying various forms of such hate. The systematic creation of a meticulously annotated Arabic hate speech dataset, specifically focused on women's sports, enhances the dataset's reliability and provides valuable insights for future research in countering hate speech against women in sports. This dataset forms a strong foundation for developing effective strategies to address online hate within the unique context of women's sports. The research findings contribute to the ongoing efforts to combat hate speech against women in sports on social media, aligning with the objectives of Saudi Arabia's Vision 2030 and recognizing the significance of female participation in sports.

Hate Speech Recognition: The Role of Empathy and Awareness of Social Media Influence

RICONOSCIMENTO DELL’HATE SPEECH: IL RUOLO DELL’EMPATIA E DELLA CONSAPEVOLEZZA DELL’INFLUENZA DEI SOCIAL MEDIA Abstract Hate speech occurs within democratic societies that embrace freedom of expression and is made tangible in the social network context. It is characterized by a specific form of discrimination based on the use of verbal expressions or other media content and, usually, directed at minority groups. Although there is a lack of consensus about a unique and shared definition of hate speech, its social and personal consequences are particularly relevant for the whole society. For these reasons, it seems of crucial importance to identify hate speech recognition antecedents. The present study aimed at analyzing the relationship between hate speech recognition and specific psychological constructs, namely, empathy and awareness of social media influence. More in details, we hypothesized the association between empathy and hate speech recognition was mediated by awareness of social media influence. Data obtained from 146 participants revealed that empathy positively predicted hate speech recognition, and such relationship was mediated by awareness. Implication of such findings are discussed.

Detection of Hate-Speech Tweets Based on Deep Learning: A Review

Cybercrime, cyberbullying, and hate speech have all increased in conjunction with the use of the internet and social media. The scope of hate speech knows no bounds or organizational or individual boundaries. This disorder affects many people in diverse ways. It can be harsh, offensive, or discriminating depending on the target's gender, race, political opinions, religious intolerance, nationality, human color, disability, ethnicity, sexual orientation, or status as an immigrant. Authorities and academics are investigating new methods for identifying hate speech on social media platforms like Facebook and Twitter. This study adds to the ongoing discussion about creating safer digital spaces while balancing limiting hate speech and protecting freedom of speech. Partnerships between researchers, platform developers, and communities are crucial in creating efficient and ethical content moderation systems on Twitter and other social media sites. For this reason, multiple methodologies, models, and algorithms are employed. This study presents a thorough analysis of hate speech in numerous research publications. Each article has been thoroughly examined, including evaluating the algorithms or methodologies used, databases, classification techniques, and the findings achieved. In addition, comprehensive discussions were held on all the examined papers, explicitly focusing on consuming deep learning techniques to detect hate speech.

Fighting Hate and Hate Speech: Raising Anti-Hate Awareness through Critical Analysis of Popular Cultural Texts on an Undergraduate Course

Central to the understanding of hate is an apprehension of the complexities of various hate-motivated social attitudes, which include Othering and the production of social, economic, and political hierarchies of domination. While hate speech is endemic both online and offline in contemporary society, Korean youths have difficulties recognizing its structural forces. The present study aims to offer an instructional model of a college-level course for identifying and countering hate in everyday life. As participants in this course, students read popular cultural texts thematizing hate, wrote critical reviews, and held group discussions to develop anti-hate critical thinking and raise awareness about online hate speech and hate-motivated social behavior. They showed significant progress in the surveys, which measured anti-hate critical thinking, as well as during the course, as they proceeded from observing and identifying hate speech to formulating and articulating proactive strategies to challenge it. This study provided an opportunity for college students to develop good citizenship in reading hate speech and representations of hate in popular cultural texts with a critical eye, and to reflect on the problem of hate in society.

Disinformation and Hate Speech on Social Media, Data Protection in the European Union

The large scale of the rise of hate speech on social media and disinformation by a large mass of social media users push governments and states to influence and take preventive measures. Today, social media is used, and in many cases through which hate speech is increasing. Given the impact on society, the spread of hate speech represents the main component that also marks an increase in different spheres. The Member States of the European Union are the basis for preventing and sanctioning hate speech online and have the Charter of Fundamental Rights of the European Union. But the fact that it results in a widespread practice for the European population: threats, violence, blackmail, disinformation, and intolerable language through social media, raised on various grounds such as race, ethnic background and sexual orientation. Disinformation, hate speech on social media and data protection, except for being regulated at the central level, as such, to deal with cases relating to these issues, Member States at state level have taken additional measures to prevent cases and to influence the reduction of victims generated by social media. The finding is marked by a sufficient number of victims through social media with multiple users of hate speech. The number of users identifying hate speech has decreased year-on-year.

Transfer learning for hate speech detection in social media

Today, the internet is an integral part of our daily lives, enabling people to be more connected than ever before. However, this greater connectivity and access to information increase exposure to harmful content, such as cyber-bullying and cyber-hatred. Models based on machine learning and natural language offer a way to make online platforms safer by identifying hate speech in web text autonomously. However, the main difficulty is annotating a sufficiently large number of examples to train these models. This paper uses a transfer learning technique to leverage two independent datasets jointly and builds a single representation of hate speech. We build an interpretable two-dimensional visualization tool of the constructed hate speech representation—dubbed the Map of Hate—in which multiple datasets can be projected and comparatively analyzed. The hateful content is annotated differently across the two datasets (racist and sexist in one dataset, hateful and offensive in another). However, the common representation successfully projects the harmless class of both datasets into the same space and can be used to uncover labeling errors (false positives). We also show that the joint representation boosts prediction performances when only a limited amount of supervision is available. These methods and insights hold the potential for safer social media and reduce the need to expose human moderators and annotators to distressing online messaging.