Abusive Comments Research Articles

Social Media Response to the Introduction of the Swallowable Gastric Balloon Treatment for Severe Obesity in the NHS: A Snapshot of Public Opinion.

Swallowable gastric balloons are available in the UK to treat severe obesity. Our hospital introduced this treatment in 2023, the first to do so in the National Health Service (NHS). The event was featured by the British Broadcasting Corporation (BBC) on regional television, in multiple radio interviews, and online by numerous news outlets in February 2024. Obesity and the treatment of obesity are often subject to stigma, prejudice and bias. The event gave an opportunity to observe public opinion in response to a new publicly available obesity treatment. Qualitative review of public comments in articles responding to the BBC story about the introduction of swallowable gastric balloon therapy in the NHS. Comments were categorized as positive, neutral, or negative. Out of 2364 comments reviewed from all sources, 16.6% were positive, 48.9% were neutral and 34.8% were negative. Obesity stigma was highly prevalent in the responses and included many derogatory and abusive comments, including towards the patients featured in the BBC articles. Obesity stigma is highly prevalent in those responding by social media and on news websites to a new treatment within the NHS. Negative stereotypes may be a barrier to obesity treatment within the NHS and need to be addressed.

Part of the Job? The Effect of Exposure to the Online Intimidation of Politicians on Political Ambition

It is increasingly common for politicians to experience online harassment. Even though numerous studies show the prevalence and impact of online abuse towards politicians, the broader societal impact of this phenomenon is understudied. This study focuses on the deterring effect that witnessing online abuse towards politicians could have on bystanders, testing to what extent online intimidation influences the appeal of political office. This pre-registered project examines whether exposure to abusive or sexist Twitter comments towards politicians weakens the interest of UK citizens, women especially, in pursuing political office. We hypothesized that seeing intimidating Twitter comments addressed to politicians leads to less political ambition in bystanders, especially if their gender identity aligns with that of the targeted politician. Sexist abusive comments were expected to have a greater impact on women participants than non-gendered abusive comments. The findings show that citizens are generally not easily deterred by exposure to abuse. However, we find differential effects that are concerning for political participation and female representation. There is evidence that sexist comments do discourage women observers. Respondents also generally espouse very negative expectations about the online experience of political aspirants.

Postdigital Bodies: Young People’s Experiences of Algorithmic, Tech-Facilitated Body Shaming and Image-Based Sexual Abuse during and after the COVID-19 Pandemic in England

In this paper, we draw upon a study exploring how COVID-19 and social isolation impacted young people’s (aged 13–18) experiences of online sexual and gendered risks and harms in England during nationwide lockdowns and upon their return to school. We explore the complexities, tensions and ambiguities in youth navigating algorithmised feeds on social media apps such as TikTok and content featuring idealised cis-gendered, heterosexualised feminine and masculine embodiment. Young people repeatedly witness hateful and abusive comments that are algorithmically boosted. We argue that this toxic content normalises online hate in the form of body shaming and sexual shaming, developing the concept of the postdigital to analyse the offline, affective, embodied and material dimensions of online harm, harassment and abuse. We also explore young people’s direct experiences of receiving harmful comments, including girls’ and gender and sexuality-diverse youth’s experiences of body and sexual shaming, as well as boys’ experiences of fat shaming; which, in many instances, we argue must be classified as forms of image-based abuse. Using our postdigital lens, we argue that the ways heteronormative, cis-gendered masculine and feminine embodiment are policed online shapes behaviour and norms in young people’s everyday lives, including in and around school, and that better understanding and support around these issues is urgently needed.

Analisis Isi Video Di Media Sosial Channel Youtube Kinderflix Dalam Kasus Pelecehan

The development of information technology, especially through social media platforms such as YouTube, has rapidly influenced the dynamics of social life in Indonesia. However, along with this growth comes new challenges, especially regarding harassment and negative comments on these platforms. This research takes a case study of the YouTube channel Kinderflix, which provides educational content for children, to explore the impact of abusive comments on content creators. The research results show that negative comments not only have a significant emotional but also psychological impact, especially on content creators. Protection of content creators and more effective management of content on social media platforms is becoming increasingly important. In the context of children's content, the protection of content creators and the welfare of children must be a top priority.

Abusive Comments (Hate Speech) on Indonesian Social Media: A Forensic Linguistics Approach

Online hate speech has significantly increased, particularly on social media, as a result of the internet's democratization of communication. This is an excellent illustration of how technology has both benefits and drawbacks. The investigation seeks to discover abusive comments on Indonesian social media. The goal of this study is to use a forensic linguistics approach to look at abusive comments on Indonesian social media (Instagram) that could lead to legal offenses. This study employs descriptive qualitative research methods. The study used the speech act theory of Austin (1962) and Searle (1969). The utterances are examined and evaluated within the context of Indonesian legal regulations. The data was obtained from @lambe_turah account (2022). The analysis's findings show that hate speech expresses negative attitudes. This can be seen in acts of insults (expressive speech). Based on the research's findings, it can be concluded that abusive comments are implicit in nature and contain implied meanings. Furthermore, there was a criminal act committed in the comments shared on Indonesian social media platforms, which involved the intentional expression of animosity towards a specific individual or group. This statement has the potential to spark communal anger and frustration because it is presented publicly and can be viewed by a large number of people. The exhaustive analysis contributes to a deeper comprehension of the potential legal implications of hate speech on social media. Additionally, the significance of courteous language is emphasized through adherence to established norms and laws.

Abusive Comment Detection in Tamil Code-Mixed Data by Adjusting Class Weights and Refining Features

In recent years, a significant portion of the content on various platforms on the internet has been found to be offensive or abusive. Abusive comment detection can go a long way in preventing internet users from facing the adverse effects of coming in contact with abusive language. This problem is particularly challenging when the comments are found in low-resource languages like Tamil or Tamil-English code-mixed text. So far, there has not been any substantial work on abusive comment detection using imbalanced datasets. Furthermore, significant work has not been performed, especially for Tamil code-mixed data, that involves analysing the dataset for classification and accordingly creating a custom vocabulary for preprocessing. This paper proposes a novel approach to classify abusive comments from an imbalanced dataset using a customised training vocabulary and a combination of statistical feature selection with language-agnostic feature selection while making use of explainable AI for feature refinement. Our model achieved an accuracy of 74% and a macro F1-score of 0.46.

Enhancing User Security on Instagram: A Multifaceted AI System for Filtering Abusive Comments

Enhancing User Security on Instagram: A Multifaceted AI System for Filtering Abusive Comments

Toxic Comment Classification using Deep Learning

Online Conversation media serves as a means for individuals to engage, cooperate, and exchange ideas; however, it is also considered a platform that facilitates the spread of hateful and offensive comments, which could significantly impact one's emotional and mental health. The rapid growth of online communication makes it impractical to manually identify and filter out hateful tweets. Consequently, there is a pressing need for a method or strategy to eliminate toxic and abusive comments and ensure the safety and cleanliness of social media platforms. Utilizing LSTM, Character-level CNN, Word-level CNN, and Hybrid model (LSTM + CNN) in this toxicity analysis is to classify comments and identify the different types of toxic classes by means of a comparative analysis of various models. The neural network models utilized for this analysis take in comments extracted from online platforms, including both toxic and non-toxic comments. The results of this study can contribute towards the development of a web interface that enables the identification of toxic and hateful comments within a given sentence or phrase, and categorizes them into their respective toxicity classes.

User-aware multilingual abusive content detection in social media

Despite growing efforts to halt distasteful content on social media, multilingualism has added a new dimension to this problem. The scarcity of resources makes the challenge even greater when it comes to low-resource languages. This work focuses on providing a novel method for abusive content detection in multiple low-resource Indic languages. Our observation indicates that a post’s tendency to attract abusive comments, as well as features such as user history and social context, significantly aid in the detection of abusive content. The proposed method first learns social and text context features in two separate modules. The integrated representation from these modules is learned and used for the final prediction. To evaluate the performance of our method against different classical and state-of-the-art methods, we have performed extensive experiments on SCIDN and MACI datasets consisting of 1.5M and 665K multilingual comments, respectively. Our proposed method outperforms state-of-the-art baseline methods with an average increase of 4.08% and 9.52% in the F1-score on SCIDN and MACI datasets, respectively.

Detecting homophobic and transphobic texts from youtube comments using machine learning models

Today's world has witnessed an exponential rise in disseminating degrading and offensive content via social media. A global increase in violence against minorities, such as gun violence, murders, and forced displacement, has been connected to using harsh and derogatory language online. The policies enacted to prevent abusive or derogatory language risk stifling free speech and are applied differently. These languages can affect the mental state of social media users. Homophobic and transphobic expressions are a insulting people's sexuality or character among harsh and disrespectful remarks. To study social media information and discriminate between homophobic and transphobic comments, it is necessary to construct language-based automatic categorization methods. This dataset consists of Tamil YouTube comments collected as part of the Shared Task on Sentiment Analysis and Homophobia detection of YouTube comments in Code-Mixed Dravidian Languages. This study detects and categorises abusive comments using machine learning models such as SVM, Naive Bayes, Logistic regression, and KNN. In comparison to other classifiers, the accuracy of logistic regression is the highest, at 55%.

‘What a lying slut’: the (re)production of rape myths in online misogyny towards women disclosing their experiences of rape through the #MeToo movement

Online misogyny is a form of online abuse against women and girls which is rooted in a hatred and mistrust of women which seeks to silence them and reinforce gender inequalities. This study extends existing research in this area by examining online misogyny within the context of the #MeToo movement. The #MeToo movement created a safe space for women to share their experiences of sexual violence, and while this promoted solidarity, support and healing, there was also an abusive backlash towards the women involved. Based on a thematic analysis of abusive YouTube comments directed at women sharing their experiences of rape, it was found that 69 per cent of abusive comments were characterised by one or more of the seven rape myths: she asked for it, it wasn’t really rape, he didn’t mean to, she wanted it, she lied, rape is a trivial event, and rape is a deviant event. This article discusses each rape myth and demonstrates how rape myths are (re)produced in online misogyny towards victims/survivors of rape and concludes by arguing this response attempts to discredit victims/survivors, the #MeToo movement and disrupt the safe space the movement created to derail the collective conversation about men’s violence against women and girls.

Netnographic Study of Online Gender-Based Violence (KBGO) on Twitter

This study aims to determine online gender-based violence in the form of online sexual harassment (Cyber Sexual Harassment) that occurs on Twitter. This research uses CMC (Computer-Mediated Communication) theory, and the type of research used is qualitative research using ethnographic methods. The results of this study show that there are six categories of cyber sexual harassment, the first is a direct threat of sexual or physical violence which is any effort, either physical or verbal that leads to a person's sexuality or sexual organs without obtaining consent and has elements or threats. Second, online harassment is an act of sexual harassment through unwanted messages, attention, and contact or intended to embarrass the victim. Third, abusive comments are negative words and without censorship can even be read or may be shared by others.

Detecting abusive comments at a fine-grained level in a low-resource language

YouTube is a video-sharing and social media platform where users create profiles and share videos for their followers to view, like, and comment on. Abusive comments on videos or replies to other comments may be offensive and detrimental for the mental health of users on the platform. It is observed that often the language used in these comments is informal and does not necessarily adhere to the formal syntactic and lexical structure of the language. Therefore, creating a rule-based system for filtering out abusive comments is challenging. This article introduces four datasets of abusive comments in Tamil and code-mixed Tamil–English extracted from YouTube. Comment-level annotation has been carried out for each dataset by assigning polarities to the comments. We hope these datasets can be used to train effective machine learning-based comment filters for these languages by mitigating the challenges associated with rule-based systems. In order to establish baselines on these datasets, we have carried out experiments with various machine learning classifiers and reported the results using F1-score, precision, and recall. Furthermore, we have employed a t-test to analyze the statistical significance of the results generated by the machine learning classifiers. Furthermore, we have employed SHAP values to analyze and explain the results generated by the machine learning classifiers. The primary contribution of this paper is the construction of a publicly accessible dataset of social media messages annotated with a fine-grained abusive speech in the low-resource Tamil language. Overall, we discovered that MURIL performed well on the binary abusive comment detection task, showing the applicability of multilingual transformers for this work. Nonetheless, a fine-grained annotation for Fine-grained abusive comment detection resulted in a significantly lower number of samples per class, and classical machine learning models outperformed deep learning models, which require extensive training datasets, on this challenge. According to our knowledge, this was the first Tamil-language study on FGACD focused on diverse ethnicities. The methodology for detecting abusive messages described in this work may aid in the creation of comment filters for other under-resourced languages on social media.

Leveraging posts’ and authors’ metadata to spot several forms of abusive comments in Twitter

Leveraging posts’ and authors’ metadata to spot several forms of abusive comments in Twitter

Abusive comment identification on Indonesian social media data using hybrid deep learning

<span lang="EN-US">Half of the entire social media users in Indonesia has experienced cyberbullying. Cyberbullying is one of the treatments received as an attack with abusive words. An abusive word is a word or phrase that contained harassment and is expressed be it spoken or in the form of text. This is a serious problem that must be controlled because the act has an impact on the victim's psychology and causes trauma resulting in depression. This study proposed to identify abusive comments from social media in Indonesian language using a deep learning approach. The architecture used is a hybrid model, a combination between <a name="_Hlk99537661"></a>recurrent neural network (RNN) and long short-term memory (LSTM). RNN can map the input sequences to fixed-size vectors on hidden vector components and LSTM implemented to overcome gradient vector growth components that have the potential to exist in RNN. The steps carried out include preprocessing, modelling, implementation, and evaluation. The dataset used is indonesian abusive and hate speech from Twitter data. The evaluation result showed that the model proposed produced an f-measure value of 94% with an increase in accuracy of 23%.</span>

KRAYA: Shop Management and Auto Services Portal

Abstract: Online shopping and travel has influenced the way we live in today’s world. It has made life comfortable and easier. The growth of these have completely changed the shopping and travel scenario as the customer does not have to physically visita shop or wait for long to avail a travel service. This has affected the shops and travel options like auto rickshaws in rural areas. The over reliability of these services has reduced the sales and income of small scale shops and auto rickshaw drivers. The main objective of this web application is to highlight the services in rural areas and provide a platform for people to access these services easily. This web portal brings together shops and auto drivers in rural areas to showcase their products and improve their income. Here, the user can find information about various shops and their products and shopkeepers can add products and sell them. The feature of auto rickshaw availability provides information about auto rickshaws that are available for travel and we can access them according to our requirements. A chat bot feature is also provided for addressing queries and filtering abusive comments. The main emphasis lies in providing a user- friendly experience where the user gets the desired results specificto their needs. Index Terms: shop, auto service, rural areas, chatbot.

Online Shaming on Social Media: Analyze and Mitigation

In this paper we would be discussing about online shaming and analyse different type of shaming and how to deal with it. The task of public shaming detection in social media is automated from the perspective of victims. It explores primarily about two aspects, namely, events and shamers. Based on classification of shaming tweets, a web application has been designed and deployed especially for one type of shaming tweet that of sarcasm, and the website also provide information about shamer who has used abusive comments under the user profile more than three time and sent alert message to user about the informing about the shamer

First, do no harm: automated detection of abusive comments in student evaluation of teaching surveys

ABSTRACT Student evaluation of teaching (SET) surveys are the most widely used tool for collecting higher education student feedback to inform academic quality improvement, promotion and recruitment processes. Malicious and abusive student comments in SET surveys have the potential to harm the wellbeing and career prospects of academics. Despite much literature highlighting abusive feedback in SET surveys, little research attention has been given to methods for screening student comments to identify and remove those that may cause harm to academics. This project applied innovative machine learning techniques, along with a dictionary of keywords to screen more than 100,000 student comments made via a university SET during 2021. The study concluded that these methods, when used in conjunction with a final stage of human checking, are an effective and practicable means of screening student comments. Higher education institutions have an obligation to balance the rights of students to provide feedback on their learning experience with a duty to protect academics from harm by pre-screening student comments before releasing SET results to academics.

DETECTING ABUSIVE AND INSULTING COMMENTS ON SOCIAL MEDIA

Now a days, Social media has become a very important mode of communication. Billions of people communicate through various social networking sites. People are using these sites as a medium to express their views and opinions. As the faces of the coin, there are both pros and cons by using this medium. One of the major problems is the insulting and the abusive comments that we get on the social media. Hence, the need for good quality automated abusive language classifiers becomes important. Our project “Detecting Abusive and Insulting Comments on Social Media” aims at designing a system for automatically detecting comments on social media as abusive or non-abusive using machine learning algorithms such as Random Forest classifier and Artificial Neural Network techniques (ANN).

Abusive comments in student evaluations of courses and teaching: the attacks women and marginalised academics endure

This paper examines the volume and type of anonymous comments academics receive in student evaluations of courses and teaching (SETs) at the 16,000 higher education institutions that collect this data at the end of each teaching period. Existing research has increasingly pointed to the negative issues of student surveys, but very little research has focused on the volume, type, and impact of anonymous student comments on academics. This paper analyses the survey results of 674 academics to inform higher education leaders and the sector more widely of the amount and type of abusive comments academics are receiving. The work also demonstrates that the highest volume, most derogatory, and most threatening abuse is directed towards women academics and those academics from marginalised groups. The paper finds that previous estimates of the rate and severity of abusive comments that academics receive, and the impact to academics’ wellbeing, mental health, and career progression, have underestimated what is taking place. The paper argues that many universities are failing to protect their staff from this abuse, and the prejudice nature of SET results, which will continue to have a negative impact on the career progression of marginalised academics - a major flaw in a sector that prides itself on diversity and inclusion.