Cyberbullying on social media under the influence of COVID‐19

Machine learning and deep learning-based approach to categorize Bengali comments on social networks using fused dataset

Through the advancement of the contemporary web and the rapid adoption of social media platforms such as YouTube, Twitter, and Facebook, for example, life has become much easier when dealing with certain highly personal problems. The far-reaching consequences of online harassment require immediate preventative steps to safeguard psychological wellness and scholarly achievement via detection at an earlier stage. This piece of writing aims to eliminate online harassment and create a criticism-free online environment. In the paper, we have used a variety of attributes to evaluate a large number of Bengali comments. We communicate cleansed data utilizing machine learning (ML) methods and natural language processing techniques, which must be followed using term frequency and reverse document frequency (TF-IDF) with a count vectorizer. In addition, we used tokenization with padding to feed our deep learning (DL) models. Using mathematical visualization and natural language processing, online bullying could be detected quickly. Multi-layer Perceptron (MLP), K-Nearest Neighbors (K-NN), Extreme Gradient Boosting (XGBoost), Adaptive Boosting Classifier (AdaBoost), Logistic Regression Classifier (LR), Random Forest Classifier (RF), Bagging Classifier, Stochastic Gradient Descent (SGD), Voting Classifier, and Stacking are employed in the research we conducted. We expanded our investigation to include different DL frameworks. Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), Convolutional-Long Short-Term Memory (C-LSTM), and Bidirectional Long Short-Term Memory (BiLSTM) are all implemented. A large amount of data is required to precisely recognize harassing behavior. To rapidly recognize internet harassment written material, we combined two sets of data, producing 94,000 Bengali comments from different points of view. After understanding the ML and DL models, we can see that a hybrid model (MLP+SGD+LR) performed more effectively when compared to other models, its evaluation accuracy is 99.34%, precision is 99.34%, recall rate is 99.33%, and F1 score is 99.34% on multi-label class. For the binary classification model, we got 99.41% of accuracy.

Does the COVID-19 Pandemic Increase or Decrease the Global Cyberbullying Behaviors? A Systematic Review and Meta-Analysis

Although cyberbullying is an emerging public health problem, it is unclear how the COVID-19 pandemic affects cyberbullying. This systematic review and meta-analysis aimed to examine the effect of the COVID-19 pandemic on cyberbullying, to estimate the global cyberbullying prevalence and to explore factors related to cyberbullying during the COVID-19 pandemic. We searched the Medline, Embase, PubMed, Scopus, Eric, PsycINFO, Web of Science, Cochrane Library, Wanfang, Chinese CNKI, and EBSCO databases to identify relevant empirical studies published between 2019 and 2022. A total of 36 studies were included. Quality assessment, meta-analyses, and subgroup analyses were conducted. The pooled prevalences were 16% for overall cyberbullying, 18% for victimization and 11% for perpetration during the COVID-19 pandemic, which were lower than before the COVID-19 pandemic. The pooled prevalence of postpandemic cyberbullying perpetration is lower in children than in adults. In addition, both virus- and lockdown-related stressors were the main factors contributing to cyberbullying. The COVID-19 crisis may reduce cyberbullying, and the pooled prevalence of cyberbullying during the pandemic in adults is higher than in children and adolescents. In addition, the transient-enduring factor model of postpandemic cyberbullying built in this review could help identify people at high risk of cyberbullying during public health emergencies.

Unmasking the Pandemic's Hidden Challenge

In the COVID-19 pandemic, a new challenge emerges in the digital realm – a surge in cyberbullying fuelled by social isolation. This chapter explores the intricate connection between isolation measures and the alarming rise in online harassment. Lockdowns and distancing have turned the online space into both a refuge and a battlefield, revealing unique consequences of isolation. The pandemic has reshaped cyberbullying, with increased screen time, online education, and heightened social media engagement playing complex roles. As social isolation heightens vulnerability, the line between physical and digital blurs, exposing individuals to online harassers. The chapter acknowledges the profound impact of social isolation on mental health, with pandemic stressors and online anonymity fostering anxiety among cyberbullying victims. Educational settings, grappling with remote learning challenges, also experience the ripple effects of cyberbullying. Despite these challenges, coping strategies and support mechanisms emerge, aiding individuals in navigating the digital landscape safely.

Has the COVID-19 Pandemic Affected Cyberbullying and Cybervictimization Prevalence among Children and Adolescents? A Systematic Review

In light of the alarming results emerging from some studies and reports on the significant increase in aggressive online behaviors among children and adolescents during the COVID-19 pandemic, the current research aimed at providing a more detailed evaluation of the investigations focusing on the cyberbullying prevalence rates published between 2020 and 2023. To this purpose, systematic searches were conducted on four databases (Web of Science, APA PsycInfo, Scopus and Google Scholar), and following PRISMA guidelines, 16 studies were included and qualitatively reviewed. Although studies were characterized by a large variety in cyberbullying operationalization and measurement, and by different methodologies used for data collection, the prevalence rates of the involvement in cyberbullying and/or cybervictimization generally revealed opposite trends: an increase in many Asian countries and Australia and a decrease in Western countries. The findings were also discussed by considering the effects of the COVID-19 pandemic. Finally, some suggestions were provided to policy makers for promoting prevention and intervention anti-cyberbullying programs in school contexts.

Clinical and Forensic Aspects of the Management of Child Abuse: The Experience of the Paediatric Emergency Department in Novara, North-West Italy

BACKGROUND

Child abuse is an endemic phenomenon that refers to any form of violence aimed at children and adolescents. The Emergency Room is often the entry point to healthcare for the abused child.

METHODS

This is a cross-sectional study including minors, aged 0-18 years, of all genders, who experienced any form of violence examined at the Paediatric Emergency Department of the 'Maggiore della Carità' Hospital in Novara (North-West Italy) between 1 January 2017 to 31 December 2021. Data were extrapolated by looking at the diagnosis at discharge. A comparison of the different variables collected was made between the pre-COVID-19 period and the COVID era.

RESULTS

120 minors presented to the paediatric emergency room seeking help for violence. The average age was 10 years, 55% of the victims were male and 75% of them were Italian. In the pre-COVID period, the number of presentations for abuse was 62, while in the COVID period it was 58 with an increase of peer violence (from 38.71% to 62.07%) and with a statistically significant impact of the pandemic on the phenomenon (p-value < 0.00001). In general, peer violence accounts for 50% of the cases reviewed and resulted in fewer reports to the judicial authority and requests for forensic advice.

CONCLUSION

The SARS-CoV-2-related pandemic has had an impact on total emergency room admissions and the types of abuse perpetrated.

Epidemiology of peer cybervictimization and its relationship with health‐related quality of life in adolescents: A prospective study

INTRODUCTION

Research focused on the association between peer cybervictimization and declining health-related quality of life (HRQoL) is scarce. Currently, few longitudinal studies find an association between these phenomena, and none focus on cybervictimization profiles. The main objectives are: (1) to analyze the point and period prevalence, and incidence of cybervictimization profiles (uninvolved, new, ceased, intermittent, and stable cybervictims); (2) to study the relationship between cybervictimization and HRQoL over time; (3) to determine the longitudinal impact on the HRQoL of each type of profile.

METHODS

A prospective study was conducted in three waves over 13 months. A total of 1142 adolescents aged 11-18 years participated in all the waves (630 girls, 55.2%).

RESULTS

The prevalence of victimization for the three waves was 21.6% (Wave 1; W1), 23.5% (W2), and 19.6% (W3), respectively. The period prevalence was 41.3%, and the accumulated incidence was 25.1%. It was found that 24% of the participants were new victims, 5.9% were intermittent victims, and 6% were stable victims. Being a cybervictim at W1 poses a relative risk of 1.73 [1.29-2.32], that is, a twofold increased risk of presenting a low HRQoL 13 months later compared to those who are not cybervictims.

CONCLUSION

One in four adolescents became a new cybervictim during the 13 months of the study. The adolescents who presented poorer HRQoL were the stable cybervictims.