Harmful Online Research Articles

Online disinhibition mediates the relationship between emotion regulation difficulties and uncivil communication

Uncivil online communication is a widely problematized cultural by-product of computer-mediated communication that, however, remains theoretically underexplained. While previous research shows that personal tendency for uncivil communication is partially influenced by individuals’ personality and empathy skills, the factors of inter-individual variation remain largely unknown. The present study examined individuals’ emotion regulation skills as a possible predictor of uncivil communication. Online survey respondents (N = 215) reported if they had engaged in uncivil communication and filled in scales measuring emotion regulation difficulties, use of different emotion regulation strategies, and various individual traits. The results show that emotion regulation difficulties were associated with high levels of online disinhibition. This, in turn, was associated with reports of uncivil communication. The mediation effect was observed even when controlling for personality and empathy. The results suggest that individuals’ emotion regulation difficulties may be an underlying psychological factor behind harmful online communication. These findings call for research and development of means to support emotion regulation in online interactions.

The UK Online Safety Act, the EU Digital Services Act and online disinformation: is the right to political participation adequately protected?*

ABSTRACT Online disinformation is a constant on social media platforms, particularly flourishing in times of elections. While sometimes harmless, it can lead to very serious human rights violations, including the right to political participation. The recent Online Safety Act in the United Kingdom and the EU Digital Services Act promise to make the online space safe by addressing harmful online content. This article explores the extent to which the OSA and DSA protect the right to political participation when addressing online disinformation. By comparing the two instruments, the article argues that they fall short in effectively protecting the two main components of Article 25 ICCPR, the right to vote and the right to stand for election, when threatened by online disinformation. In particular, it shows that the existing provisions are often vague and general with excessive deference to online platforms, urging strong enforcement and a more prominent role for regulators.

Knowledge, deception, and freedom of expression: a critical examination of Ireland’s approach to disinformation under the Electoral Reform Act 2022

ABSTRACT Ireland’s Electoral Reform Act introduces several key provisions under Parts 4 and 5 which are designed to curtail the spread of misleading electoral communications. This article’s focus is on how these provisions enable divergence from established EU standards surrounding intermediary liability for harmful online communications. Identifying how the application of these provisions may undermine the right to freedom of expression under EU law, this article further distils the applicable principles which should inform how online platforms limit access to misleading information in electoral settings. This article filters these principles from applicable case law of the Court of Justice of the European Union (CJEU) and the European Court of Human Rights (ECtHR). Drawing from the analytical approaches of these courts to false electoral communications, this article not only assesses how Ireland’s Electoral Reform Act may undermine the right to freedom of expression but also provides a template for how EU and Member State legislation in the disinformation field can promote compatibility with human rights.

The role of digital cues in online banter: a systematic review of adolescents’ and emerging adults’ views, uses and interpretations

The potential ambiguity and aggression related to banter-type humorous interactions may lead to misinterpretations and escalations such as perceived bullying, especially in the online context. A way to minimise such risks during online banter is the use of digital visual cues such as emojis, emoticons, GIFs, personal photos, videos, etc. This review summarised and synthesised the literature to understand how adolescents and emerging adults perceive and interpret the role of digital visual cues in online banter/humour and how they actually use these cues in banter-type online interactions. From the 12,372 papers identified, 17 papers were included, and the relevant narrative data were analysed using Reflexive Thematic Analysis (Clarke & Braun, 2021). Four themes were generated, including, (1) Digital cues are an essential part of digital banter, (2) Digital cues are ambiguous, (3) Harmful online banter containing digital cues, and (4) Personal and contextual effects. Implications and future directions are discussed.

The Dilemma and Solutions of the New Phenomena in the Network Subculture of Generation Z

As a new force on the internet, Generation Z has created a landscape of network subcultures in the new media era. They gather in groups based on shared interests, forming circle cultures such as the lying flat culture, partner culture, MBTI culture, and mystery box culture. Additionally, through the collage, deconstruction, and reconstruction of linguistic symbols, new internet terms like crazy literature, rat literature, and abbreviations have emerged. However, these emerging network subcultures face dilemmas such as label solidification due to excessive circling, bottomless traffic hype driven by entertainment supremacy, and value distortions caused by a lack of regulatory measures. To address these issues, Generation Z need to improve their information discernment capabilities to identify harmful online content. Mainstream media should guide appropriately to maintain the innovation and vitality of network subcultures. Meanwhile, the government should focus on the physical and mental health and needs of Generation Z, actively guiding them in scientifically relieving stress.

Alternative cancer clinics’ use of Google listings and reviews to mislead potential patients

BackgroundAlternative cancer clinics, who provide treatment associated with earlier time to death, actively seek to create favorable views of their services online. An unexplored means where alternative cancer clinics can shape their appeal is their Google search results.MethodsWe retrieved the Google listing and Google reviews of 47 prominent alternative cancer clinics on August 22, 2022. We then conducted a content analysis to assess the information cancer patients are faced with online.ResultsGoogle listings of alternative treatment providers rarely declared the clinic was an alternative clinic versus a conventional primary cancer treatment provider (12.8% declared; 83.0% undeclared). The clinics were highly rated (median, 4.5 stars of 5). Reasons for positive reviews included treatment quality (n = 519), care (n = 420), and outcomes (n = 316). 288 reviews presented the clinics to cure or improve cancer. Negative reviews presented alternative clinics to financially exploit patients with ineffective treatment (n = 98), worsen patients’ condition (n = 72), provide poor care (n = 41), and misrepresent outcomes (n = 23).ConclusionsThe favorable Google listing and reviews of alternative clinics contribute to harmful online ecosystems. Reviews provide compelling narratives but are an ineffective indicator of treatment outcomes. Google lacks safeguards for truthful reviews and should not be used for medical decision-making.

Generalizing Hate Speech Detection Using Multi-Task Learning: A Case Study of Political Public Figures

Automatic identification of hateful and abusive content is vital in combating the spread of harmful online content and its damaging effects. Most existing works evaluate models by examining the generalization error on train–test splits on hate speech datasets. These datasets often differ in their definitions and labeling criteria, leading to poor generalization performance when predicting across new domains and datasets. This work proposes a new Multi-task Learning (MTL) pipeline that trains simultaneously across multiple hate speech datasets to construct a more encompassing classification model. Using a dataset-level leave-one-out evaluation (designating a dataset for testing and jointly training on all others), we trial the MTL detection on new, previously unseen datasets. Our results consistently outperform a large sample of existing work. We show strong results when examining the generalization error in train–test splits and substantial improvements when predicting on previously unseen datasets. Furthermore, we assemble a novel dataset, dubbed PubFigs, focusing on the problematic speech of American Public Political Figures. We crowdsource-label using Amazon MTurk more than 20,000 tweets and machine-label problematic speech in all the 305,235 tweets in PubFigs. We find that the abusive and hate tweeting mainly originates from right-leaning figures and relates to six topics, including Islam, women, ethnicity, and immigrants. We show that MTL builds embeddings that can simultaneously separate abusive from hate speech, and identify its topics.

A Comprehensive Review- Building A Secure Social Media Environment for Kids- Automated Content Filtering with Biometric Feedback

This review paper is all about how important it is to use smart technology to keep kids safe on social media while helping them learn better. By adding things like better controls for parents, filters that stop bad stuff, and tools that check how kids are feeling, we can make sure they don't run into anything harmful online. In today's world where kids spend a lot of time online, it's super important to make sure they're safe. If social media platforms start using cool new tech like biometric sensors and wearable gadgets, they can create safer spaces for kids to have fun and learn. This paper also talks about why we need to do things ahead of time to deal with problems like spending too much time on screens or seeing things that might not be right for us. By giving practical ideas for researchers, people who make rules, and companies, this paper wants to make sure kids can enjoy the good parts of social media without any worries.

Mindfulness programs for problematic usage ofthe internet: A systematic review and meta-analysis.

Problematic usage of the internet (PUI) is an umbrella term for a range of uncontrolled, excessive, and potentially harmful online behaviors. Recently, numerous studies have examined the potential of mindfulness programs (MPs) for reducing PUI. We conducted a comprehensive systematic review and meta-analysis in this emerging field. We searched eight databases from inception to October 18, 2022, with no language restrictions. We included randomized controlled trials (RCTs) and nonrandomized trials (NRTs). The primary outcome was change in self-reported PUI, the secondary outcome was change in screen time. Of 3,473 identified records, 19 RCTs and 20 NRTs with a total of 1,549 participants were included. Participation in an MP was associated with large reductions in PUI in between-group analysis in RCTs (k = 19; g = -1.67; 95% CI -2.15, -1.19) and in within-group pre-post analysis in all studies (k = 35; g = -1.67; 95% CI -1.99, -1.36). Screen time showed a medium reduction in within-group pre-post analysis (k = 10; g = -0.65; 95% CI -0.90, -0.41). The effects for PUI remained significant in a series of sensitivity analyses, such as excluding low quality studies, excluding outliers, adjusting for publication bias, or using follow-up data. Heterogeneity between studies was high and the overall quality of evidence was rated low. MPs are probably effective in reducing PUI and might be effective in reducing screen time. Shortcomings in the quality of evidence highlight the need for high-quality controlled trials with long-term follow-ups to confirm results.

Sensitive and Obscene Content Blocker (SOCB)

Abstract: In today’s digital age, the internet is a vast and diverse platform that offers a wealth of information and entertainment. However, this openness also exposes users to sensitive and obscene content, which can have a detrimental impact on individuals, particularly children and vulnerable populations. To mitigate the exposure to such content, a Sensitive and Obscene Content Blocker has been developed and tested. This paper explores the design, development, and evaluation of a comprehensive content filtering system that is aimed at identifying and blocking sensitive and obscene content across various online platforms. The system employs advanced algorithms, machine learning techniques, and real-time content analysis to ensure a reliable and robust filtering mechanism. The paper outlines the key components of the blocker, including content analysis methods, their detection and classification methods. The study presents the results of extensive testing, demonstrating the system’s efficiency in blocking inappropriate content while minimizing false positives. The Sensitive and Obscene Content Blocker serves as a powerful tool to protect individuals from harmful online content, promoting a safer and more secure digital environment. This paper provides valuable insights into the development and implementation of content blockers and their role in enhancing online safety and security.

Combating Islamophobia: Compromise, Community, and Harmony in Mitigating Harmful Online Content

Despite significant advances in content moderation within HCI, social computing scholarship in this area remains constrained by secular values and Western interpretations of justice. As a result, current literature often overlooks religious and spiritual sensibilities, as well as communal peacebuilding efforts even when the harms originate from and strongly connected to faith sensitivities, such as Islamophobia. This article presents findings from a design and evaluation study on the reporting and moderation of Islamophobic posts on Twitter (currently known as “X”). By utilizing HCI theories and readily available NLP techniques, we developed an online tool for reporting and moderating Islamophobic tweets. We subsequently conducted usability studies, contextual inquiries, and interviews with 32 participants to assess the tool’s effectiveness in addressing Islamophobic content. Our study revealed that factors such as faith-related knowledge practices, fact-checking, communal leadership, social harmony, and the cultural-religious value of “compromise” significantly influence reactions to Islamophobic posts online. Expanding on these findings and drawing from the literature on conflict resolution in theology, legal studies, and justification, we explore how “Sulha,” a community-driven process for mitigating conflict and restoring communal peace, can cater to faith-based sensibilities in reporting and moderating Islamophobic content. Therefore, this article complements existing content moderation literature by recommending the adaptation of faith sensitivities in the design of computing tools and policies to mitigate Islamophobia and similar faith-related online harms.

Challenges in investigating self-generated online child sexual abuse material

Aim: The overall aim of this paper is to offer a greater understanding of the context of children’s online sexual behaviour through exploring what are some of the main challenges the police encounter when investigating digital content of a sexual nature produced by young people, such as photographs, videos and chat messages.Methodology: The research is primarily reviewing and analysing academic literature, policy papers, legal sources and statistical data. The main areas examined include the increasing prevalence of self-generated child sexual abuse material (CSAM), how contemporary peer-to-peer relations amongst children impact the self-production of sexual material, and how creating sexual content online for commercial reasons and the development of digital technologies add to the complexity and ambiguities of investigations.Findings: Findings in this paper highlight that law enforcement alone will not be able to manage all the challenges that online child sexual abuse crimes imply on society and children in particular. They will need clear guidelines to differentiate between children’s harmful and healthy online sexual behaviour and upto- date knowledge on how to avoid the unnecessary victimisation of children.Value: This paper promotes a more nuanced understanding of the complex phenomenon of children self-generating explicit content, which is often embedded within wider societal issues of peer-to-peer relations, financial deprivation and technological advancement. As a result, we will be able to better protect young people, distinguishing between adolescents’ healthy and harmful online

‘This is the fate of Libyan women:’ contempt, ridicule, and indifference of Seham Sergiwa

ABSTRACT As the debate on how harmful online content translates into violent offline actions continues in the Global North, this debate should be enriched by data from the Global South. In Libya, the kidnapping of a female politician (Seham Sergiwa) in 2019 was arguably foreshadowed on social media, with some analysts warning of an imminent danger to her life when tracking social media discourse in June/July 2019. Following her abduction, the outcry was loud and reverberated outside of Libya. However, this was short lived and similar occurrences happen in Libya still. This paper contributes a systematic analysis of Facebook content in the months leading up to and after Sergiwa’s disappearance. Using thematic content analysis to explore 259 posts and 4,367 comments – between January and December 2019 – this paper outlines three dominant themes: Contempt, ridicule and indifference. This study presents a refocus on the dangers of online speech while acknowledging complexities around categorising the former. The study adds nuanced understandings on online discussions in Libya and gendered dynamics in a violent, post-revolutionary environment.

Civil Liability of the Website Owner for Damages Arising from Published Harmful Content under Jordanian Law

Due to advanced Internet technology, there's a growing need for legal frameworks overseeing website creation and monitoring to prevent misuse, especially on social media. Legislators established laws holding website owners responsible for content under their supervision, requiring them to prevent and remove offensive material. Victims of harmful online content can seek criminal penalties, compensation, and site blocking. Like print laws, electronic publishers, and site supervisors face accountability for damaging content, prompting the need for legislation, as outlined in Jordanian Press and Publications Law Article 49. This law considers online comments as journalistic material, holding publishers and site supervisors jointly responsible. It empowers authorities to prevent harmful content publication, allowing victims to seek legal recourse in civil courts for damages.

Shifting Mindsets; A Critical Commentary on Child Protection and Welfare, Disability, and Online Risk Through Critical Literature Review

ABSTRACT Critical commentary is presented at the complex intersection of child protection and welfare, disability, and internet use. A profound lack of research exists on this topic. Findings from a critical literature review focus on the theme of online risk which included the topic of child abuse/online solicitation, cyber victimization/cyberbullying, problematic internet use, and breach of data protection. The approach taken to review literature was adopted from Aromataris and Pearson (2014). Disabled young people are exposed to and impacted by harmful online content. Key points, and online resources are highlighted for parents and guardians of disabled children, social care practitioners, safeguarding professionals, and educators. This paper attempts to advance scholarly discourse, policy, and practice, the authors arguing that it is time to shift our mindsets about safeguarding disabled young people online from thoughtful attention to purposeful action.

A Machine Learning-Based Approach for Automated Filtering and Blocking of Objectionable Web Content: Design and Implementation

This work aimed to develop a machine learning objectionable web content filtering and blocking system. This was important because of the high level of proliferation of objectionable web content which had posed a significant challenge to maintain a safe and appropriate online environment. The objectives of the study include a developed machine learning system that can send a real-time short message system(sms) to parents or guardians or other designated individuals responsible for the children’s safety informing them when their children had opened objectionable web content, a developed machine learning system that kept log of objectionable web content if network is unavailable and send notification if network is restored, a developed machine learning system endowed with enough reasoning capability to intelligently filter web objectionable content and ill-suited web pages, a developed machine learning system that was capable to have kept detailed log of objectionable web content even if the search was carried out in private (incognito mode in goggle chrome). The programming language of choice used in this work was python since its codes can be created quicker and performed faster than many other programming languages. The methodology adopted was the object-oriented Analysis and Design methodology (OOADM). The study utilized Anaconda Jupiter Notebook as its development environment, python as its programming language and then SQlite as the Database Management System (DBMS). The machine learning web content filtering and blocking system serves as a powerful tool to protect our children from harmful online content, promoting a safer and more secured digital environment.

Increasing the Effectiveness of Fake News Detection: An Educational Program for High School Students Using Interactive Neural Network Training and Collective Intelligence

This paper presents the project results designed to provide high school students with essential ICT tools to identify and counteract fake news and disinformation commonly found on the Internet, especially on platforms like X/Twitter. Additionally, it introduces an educational program that utilizes interactive neural network training and collective intelligence to combat fake news. For this project, there was developed an IT framework enabling the collective training of a specialized neural network. In order to conduct a quasi-experiment, we engaged three research groups of high school students, each containing app. 10-15 members. Through a set of comprehensive workshops, the students were trained to recognize harmful online content. After this training, students actively participated in data classification on various topics, laying the foundation for the neural network's training model. The presented results underscore the efficacy of this immersive method in imparting digital literacy and enhancing the group intelligence. Moreover, the results highlight the promising potential of machine learning in assisting youth to navigate the complex digital terrain safely and responsibly. The final phase of the conducted research involved testing the trained neural network in detecting disinformation, particularily in the topics of 5G technologies and immigration problems in Poland.

Expert-Annotated Dataset to Study Cyberbullying in Polish Language

We introduce the first dataset of harmful and offensive language collected from the Polish Internet. This dataset was meticulously curated to facilitate the exploration of harmful online phenomena such as cyberbullying and hate speech, which have exhibited a significant surge both within the Polish Internet as well as globally. The dataset was systematically collected and then annotated using two approaches. First, it was annotated by two proficient layperson volunteers, operating under the guidance of a specialist in the language of cyberbullying and hate speech. To enhance the precision of the annotations, a secondary round of annotations was carried out by a team of adept annotators with specialized long-term expertise in cyberbullying and hate speech annotations. This second phase was further overseen by an experienced annotator, acting as a super-annotator. In its initial application, the dataset was leveraged for the categorization of cyberbullying instances in the Polish language. Specifically, the dataset serves as the foundation for two distinct tasks: (1) a binary classification that segregates harmful and non-harmful messages and (2) a multi-class classification that distinguishes between two variations of harmful content (cyberbullying and hate speech), as well as a non-harmful category. Alongside the dataset itself, we also provide the models that showed satisfying classification performance. These models are made accessible for third-party use in constructing cyberbullying prevention systems.

The role of gender and sexism in the moral disengagement mechanisms of technology-facilitated sexual violence

Technology-facilitated sexual violence (TFSV) encompasses various harmful online behaviors, such as online gender-based hate speech, online gender-based violence, digital sexual harassment, image-based sexual abuse, and online sexual coercion. This study aimed to examine how moral disengagement varies based on the type of TFSV, and the factors influencing moral disengagement in TFSV situations. Specifically, the impact of participant gender, participant sexist beliefs, and victim–perpetrator gender was assessed. A total of 2950 respondents (70.3% women) between the ages of 16 and 79 (M = 26.33, SD = 10.09) participated in a vignette study in which the variables of interest (TFSV type, victim-perpetrator gender) were manipulated. The findings revealed that moral disengagement levels varied across different types of TFSV, with higher levels observed in cases of online gender-based hate speech, followed by online gender-based violence, digital sexual harassment, image-based sexual abuse, and online sexual coercion. Moreover, the results indicated that sexist beliefs, participant gender (i.e., being a man), and situations involving a woman perpetrator and a man victim were associated with greater moral disengagement. Additionally, it was found that sexist beliefs play a more significant role in moral disengagement for women than for men. This information, which emphasizes the role of gender socialization and sexist beliefs in the justification and normalization of this type of violence, is important for the development of prevention programs.

Digital Media and Younger Audiences

The active digital engagement of children and teens from a very early age makes them the most prolific digital users and online content creators. Simultaneously, this high level of digital exposure enhances their vulnerability to online risks and the potential for them to encounter harmful online content. This dynamic has profound implications for all dimensions and stakeholders within the digital ecosystem. This thematic issue presents a comprehensive review of the significant advantages, critical risks, and challenges arising from the extensive online engagement of children and adolescents. This body of research provides valuable insights and identifies future research avenues related to emotional well-being, identity development, perceptions of social success and self-esteem, as well as examining the critical aspects concerning digital literacy and the regulatory frameworks governing digital content providers.