Academic literature on the topic 'Cyberbullying representation'

In this paper, we focused on a novel method to explain cyberbullying experience among children in the Facebook. The method is Cyber Media Analysis (CMA). It is a mixed-method of analysing the stages of text and context. The text level discusses the form of messagesrelated to cyberbullying produced and received by children. They are sentences, photos or other visual representation in the Facebook that examine cyberbullying phenomenon. To obtain cyberbullying text, we conducted a content analysis of Facebook posts by children aged 8 to 12 (N=250) in the elementary schools in Solo areas. In the context level, we conducted in-depth interviews and group discussion to find out why children produced cyberbullying messages and how they reacted when receiving cyberbullying messages. Merging these two methods gives a comprehensive explanation of cyber bullying phenomenon among children.

Cyberbullying is rapidly becoming one of the most serious online risks for adolescents. This has motivated work on machine learning methods to automate the process of cyberbullying detection, which have so far mostly viewed cyberbullying as one-off incidents that occur at a single point in time. Comparatively less is known about how cyberbullying behavior occurs and evolves over time. This oversight highlights a crucial open challenge for cyberbullying-related research, given that cyberbullying is typically defined as intentional acts of aggression via electronic communication that occur repeatedly and persistently . In this article, we center our discussion on the challenge of modeling temporal patterns of cyberbullying behavior. Specifically, we investigate how temporal information within a social media session, which has an inherently hierarchical structure (e.g., words form a comment and comments form a session), can be leveraged to facilitate cyberbullying detection. Recent findings from interdisciplinary research suggest that the temporal characteristics of bullying sessions differ from those of non-bullying sessions and that the temporal information from users’ comments can improve cyberbullying detection. The proposed framework consists of three distinctive features: (1) a hierarchical structure that reflects how a social media session is formed in a bottom-up manner; (2) attention mechanisms applied at the word- and comment-level to differentiate the contributions of words and comments to the representation of a social media session; and (3) the incorporation of temporal features in modeling cyberbullying behavior at the comment-level. Quantitative and qualitative evaluations are conducted on a real-world dataset collected from Instagram, the social networking site with the highest percentage of users reporting cyberbullying experiences. Results from empirical evaluations show the significance of the proposed methods, which are tailored to capture temporal patterns of cyberbullying detection.

Despite the increasing interest in cyberbullying detection, existing efforts have largely been limited to experiments on a single platform and their generalisability across different social media platforms has received less attention. We propose XP-CB, a novel cross-platform framework based on Transformers and adversarial learning. XP-CB can enhance a Transformer leveraging unlabelled data from the source and target platforms to come up with a common representation while preventing platform-specific training. To validate our proposed framework, we experiment on cyberbullying datasets from three different platforms through six cross-platform configurations, showing its effectiveness with both BERT and RoBERTa as the underlying Transformer models.

This paper presents a novel method of analyzing morphosemantic patterns in language to the detect cyberbullying, or frequently appearing harmful messages and entries that aim to humiliate other users. The morphosemantic patterns represent a novel concept, with the assumption that analyzed elements can be perceived as a combination of morphological information, such as parts of speech, and semantic information, such as semantic roles, categories, etc. The patterns are further automatically extracted from the data containing harmful entries (cyberbullying) and non-harmful entries found on the informal websites of Japanese high schools. These website data were prepared and standardized by the Human Rights Center in Mie Prefecture, Japan. The patterns extracted in this way are further applied to a document classification task using the provided data in 10-fold cross-validation. The results indicate that morphosemantic sentence representation can be considered useful in the task of detecting the deceptive and provocative language used in cyberbullying.

This study aimed to evaluate which aspects of moral disengagement (MD), empathy, and representations of the victim’s experience (VER) could be predictors of cyberbullying (CB). One hundred and eight-nine students (11–17 years old) completed 3 self-report questionnaires: An MD scale, an empathy scale, and a CB questionnaire. In relation to the personal experience of CB, four groups were identified: Victim, bully, bully/victim, and no experience with CB. The linear bivariate correlation analysis shows correlations between empathy and VER, between empathy and MD, and between MD and VER. A multinomial logistic regression identified which predictors could increase a subject’s probability of belonging to one of the four groups regarding the personal experience of CB (victim, bully, bully/victim, no experience). Findings highlighted that low cognitive empathy might increase the probability for a student to belong to the bullies’ group, rather than the victims’ group. Furthermore, low perception of the consequences of CB on the victim might increase the probability of belonging to the bully, bully/victim, and no experience groups. Then, a high score in the diffusion of responsibility was a significant predictor of belonging to the victim group rather than the no experience group. Results from this study confirm the need for preventive measures against CB, including the empowerment of cognitive empathy, decreasing the diffusion of responsibility, and increasing the awareness of the consequences of CB on the victim.

This study was the first to examine ethnic, sex, and ethnicity-by-sex differences for underresearched, Asian American and Pacific Islander, adolescent groups on youth violence outcomes other than cyberbullying. This effort included the less researched, emotional violence, and included socioeconomic status (SES) measures as covariates. The sample size from 2 high schools in spring 2007 was 881, using an epidemiologic survey design. The pattern of results was higher rates of violence victimization for ethnic groups, with lower representation in the 2 schools’ population, and ethnic groups that more recently moved or immigrated to Hawai‘i. For emotional victimization, girls of European American and “other” ethnicities self-reported higher rates than boys. Several implications (e.g., need for ethnically and gender-based approaches) and further research (e.g., ethnocultural identity) are discussed.

Resumo Tecnologias da informação e comunicação vêm ganhando destaque como influenciadoras do desempenho cognitivo, dada a rapidez com que as informações são geradas e disseminadas, modificando relações pessoais e profissionais. Se, por um lado, isso traz evolução e transformação, por outro pode provocar danos como o cyberbullying , violência que vulnera pessoas de forma silenciosa. Essa conjuntura requer reflexão e atuação bioética. Este estudo teve como objetivo conhecer a percepção do fenômeno do cyberbullying praticado no cotidiano profissional e pessoal. Trata-se de pesquisa exploratória e quantitativa por conveniência, da qual participaram 35 servidores públicos do Instituto Federal de Educação, Ciência e Tecnologia do Sul de Minas Gerais. Os resultados foram discutidos com base na literatura sobre tecnologia da informação e comunicação, violência digital e, principalmente, bioética.

Resumen Las tecnologías de la información y la comunicación han ido ganando protagonismo como influyentes en el rendimiento cognitivo, dada la rapidez con la que se genera y difunde la información, modificando las relaciones personales y profesionales. Si, por un lado, esto trae evolución y transformación, por otro puede provocar daños como el cyberbullying , la violencia que violenta silenciosamente a las personas. Esta coyuntura requiere reflexión y actuación bioética. Este estudio tuvo como objetivo conocer la percepción del fenómeno del cyberbullying practicado en la vida cotidiana profesional y personal. Se trata de una investigación exploratoria y cuantitativa por conveniencia, en la que participaron 35 servidores públicos del Instituto Federal de Educación, Ciencia y Tecnología del Sur de Minas Gerais. Los resultados se discutieron a partir de la literatura sobre tecnología de la información y la comunicación, violencia digital y, principalmente, la bioética.

Cyberbullying is a new phenomenon, and no extensive theory has already been accredited which can describe it in a satisfactory way, although research in this field is evolving at a high speed. To gain a better understanding of the dynamics of cyberbullying it is necessary to integrate the knowledge developed in different fields of research, mainly concerned on one hand with the particularities of computer and mobile phone mediated communication and on the other hand with the fostering impact that the changes of adolescence could have on the propensity to engage in acts of cyberbullying. The idea of a virtual world whose intrinsic characteristics may enhance the probabilities of online aggressive behaviours leads to the search for variables that are valid across countries and cultures, thus to the necessity for cross-cultural research. The collaboration with the Universidade Federal Fluminense of Rio de Janeiro, the Universidad de La Sabana of Bogotà, the Middle East Technical University of Ankara as well as the support provided by the COST Action “Cyberbullying: coping with negative and enhancing positive uses of new technologies, in relationships in educational settings” and the London based organization Beatbullying.org all greatly contributed in the realization of this project, aimed at investigating the diffusion and the representation of cyberbullying among youth of different countries.

Submitted by Fernanda Ribeiro de Souza null (ferrisouza@hotmail.com) on 2018-02-09T19:16:12Z No. of bitstreams: 1 souza_fr_me_prud.pdf: 2130447 bytes, checksum: 7c77d394bf73bfaeaad3a60fbbe638b1 (MD5)
Approved for entry into archive by ALESSANDRA KUBA OSHIRO ASSUNÇÃO (alessandra@fct.unesp.br) on 2018-02-14T16:58:54Z (GMT) No. of bitstreams: 1 souza_fr_me_prud.pdf: 2130447 bytes, checksum: 7c77d394bf73bfaeaad3a60fbbe638b1 (MD5)
Made available in DSpace on 2018-02-14T16:58:54Z (GMT). No. of bitstreams: 1 souza_fr_me_prud.pdf: 2130447 bytes, checksum: 7c77d394bf73bfaeaad3a60fbbe638b1 (MD5) Previous issue date: 2017-10-24
A presente investigação, vinculada à linha de pesquisa “Práticas e Processos Formativos em Educação”, tem como embasamento o crescimento da utilização das Tecnologias Digitais de Informação e Comunicação (TDIC) entre os adolescentes, bem como das redes sociais da internet, que conduzem tanto a favoráveis situações, como podem ser utilizadas para fomentar a violência nas situações de cyberbullying. Apresenta como objetivo identificar e analisar as representações sociais de gestores, professores e alunos sobre o cyberbullying e suas manifestações nas redes sociais. Diante do objetivo geral, foram delineados os seguintes objetivos específicos: analisar as representações sociais sobre o conceito de cyberbullying por parte dos sujeitos pesquisados; identificar as representações sociais sobre o conceito de redes sociais por parte dos sujeitos pesquisados; identificar situações de cyberbullying vivenciadas pelos sujeitos pesquisados em redes sociais; analisar as representações sociais sobre o uso de redes sociais pelos adolescentes; analisar as representações sociais sobre uso de redes sociais nas atividades pedagógicas e analisar as representações sociais sobre a importância da mídia-educação como componente de formação crítica do adolescente usuário de redes sociais. A pesquisa possui abordagem qualitativa, com delineamento descritivo-explicativo. Para a análise dos dados, utilizamos a análise de conteúdo. Perante essa constatação, o trabalho foi realizado em uma escola da rede pública estadual de um município de médio porte do interior do estado de São Paulo, com a participação de estudantes, professores e gestores da escola, compreendendo como procedimentos: questionário aos estudantes e aos professores; entrevista com gestores; grupo focal com estudantes e estudo de conteúdo divulgado por meio da rede social da internet Facebook. Os resultados revelaram que, de modo geral, os participantes têm conhecimento do conceito de cyberbullying. No entanto, as situações de cyberbullying encontraram-se naturalizadas e despercebidas por parte dos participantes. No caso das redes sociais, estas foram identificadas, principalmente, como uma forma de comunicação e de relacionamento interpessoal. Seu uso na escola ainda é tímido. Acerca da mídia-educação, seu entendimento foi verificado como superficial, indicando a necessidade de ações formativas eficazes voltadas para o tema. Espera-se com a realização desse trabalho, colaborar para a reflexão e debates no âmbito educacional sobre o fenômeno do cyberbullying, além de contribuir com os sistemas educativos com sugestões sobre formas de enfrentamento e prevenção a essa violência.
The present investigation, joint to the research segment “Practice and Processes for Formation in Education”, has as base the growing process from the use of Information and Communication through Digital Technology (ICDT) among teenagers, and the social networks as well, that conduct even to favorable situations in order to increase the violence in cyber bulling situations. As an objective, it identifies and analyses the representation of the teachers, directors and students, related to cyber bullying and its manifestations at the social networks. Based on the general objective, the specific objectives were established: analyze the social representations about the concept of cyber bullying, from the researched subjects; identify the social representations about the concept of cyber bullying, from the researched subjects; identify cyber bulling experienced by the researched subjects in social networks; analyze the social representations about the use of social networks by teenagers; analyze the social representations about the use of social networks in pedagogical activities and analyze the social representations about the importance of the media and education as a component of the critical construction of the teenager as a user in social networks. The research has a qualitative approach, with descriptive and explanatory establishing. For the data analysis provided, the content analysis was used. In view of this, the work was accomplished in a State public school, from an average city, from São Paulo State, countryside, with the participation of the students, teachers and directors from the school, using the procedures: students and teachers questionnaire; interview with the directors; focus on students groups and content study, spread out through Facebook. The results showed that, in a general view, the participants have a considerable knowledge of the concept of cyber bulling. However, the situations of cyber bullying were found naturally and not really noticed by the participants. Related to the social networks, those were identified, mainly, as a way of communication and the interpersonal relationship. Its use in schools is still not common . Related to media and education, its comprehension was verified as superficial, pointing that there is a necessity of efficient forming actions focused on the theme. It is expected that, with the achievement of this research, some collaboration appear to help the reflections and debates in education itself about the phenomena cyber bullying, and also may contribute to the education systems as suggestions, about forms of facing and prevention against this kind of violence.

Since the use of social media has evolved, the ability of its users to bully others has increased. One of the prevalent forms of bullying is Cyberbullying, which occurs on the social media sites such as Facebook©, WhatsApp©, and Twitter©. The past decade has witnessed a growth in cyberbullying – is a form of bullying that occurs virtually by the use of electronic devices, such as messaging, e-mail, online gaming, social media, or through images or mails sent to a mobile. This bullying is not only limited to English language and occurs in other languages. Hence, it is of the utmost importance to detect cyberbullying in multiple languages. Since current approaches to identify cyberbullying are mostly focused on English language texts, this thesis proposes a new approach (called Multilingual Cyberbullying Detection System) for the detection of cyberbullying in multiple languages (English, Hindi, and Marathi). It uses two techniques, namely, Machine Learning-based and Lexicon-based, to classify the input data as bullying or non-bullying. The aim of this research is to not only detect cyberbullying but also provide a distributed infrastructure to detect bullying. We have developed multiple prototypes (standalone, collaborative, and cloud-based) and carried out experiments with them to detect cyberbullying on different datasets from multiple languages. The outcomes of our experiments show that the machine-learning model outperforms the lexicon-based model in all the languages. In addition, the results of our experiments show that collaboration techniques can help to improve the accuracy of a poor-performing node in the system. Finally, we show that the cloud-based configurations performed better than the local configurations.

O contínuo aparecimento do cyberbullying nas redes sociais constitui um problema mundial que tem aumentado consideravelmente nos últimos anos, e exige medidas urgentes para a deteção automática de tal fenómeno. O objetivo deste trabalho é criar um modelo suficientemente capaz de detetar automaticamente textos ofensivos. Para tal, foram utilizados três conjuntos de dados públicos, bem como duas abordagens principais para resolver este problema: uma baseada em métodos clássicos de aprendizagem automática e a outra baseada em aprendizagem profunda. Na abordagem clássica de aprendizagem automática foi proposta uma fase específica de pré-processamento e engenharia de características com várias etapas. Para além disso, foram exploradas duas abordagens de representação de documentos para gerar as entradas utilizadas pelos classificadores SVM, Logistic Regression e Random Forest. Uma vez que estes conjuntos de dados são desequilibrados, SMOTEENN e Threshold-Moving foram utilizados para lidar com o problema de classificação desbalanceada. Na abordagem de aprendizagem profunda foram exploradas diferentes arquiteturas, combinando vetores de palavras pré-treinados com CNN, CNN-Attention, BiLSTM e BiLSTM-Attention. A configuração experimental envolveu o tratamento de palavras desconhecidas, Cyclical Learning Rate para proporcionar uma melhor convergência, Macro Soft-F1 Loss para otimizar o desempenho e Macro Soft-F2 Loss para lidar com o problema de classificação desbalanceada. Foi também proposto um modelo RoBERTa-base, pré-treinado em 58 milhões de tweets e afinado para identificação de linguagem ofensiva. Os resultados experimentais mostram que, embora seja uma tarefa difícil, ambas asabordagens propostas são adequadas para detetar textos ofensivos. No entanto, a abordagem de aprendizagem profunda alcança os melhores resultados.
The continuous appearance of cyberbullying on social media constitutes a worldwide problem that has seen a considerable increase in recent years, and demands urgent measures to automatically detecting such phenomenon. The goal of this work is to create a model suficiently capable of automatically detecting ofensive texts. For this purpose, three public datasets were used, as well as two main approaches to solve this problem: one based on classical Machine Learning methods and the other based on Deep Learning. In the classical Machine Learning approach was proposed a specific pre-processing and Feature Engineering stage with several steps. In addition, two document representation approaches were also explored to generate the inputs used by SVM, Logistic Regression, and Random Forest classifiers. Since these datasets are imbalanced, SMOTEENN and Threshold-Moving were used to deal with the imbalanced classification problem. In the Deep Learning approach diferent architectures were explored, combining pretrained word vectors with CNN, CNN-Attention, BiLSTM and BiLSTM-Attention. The experimental setup involved treatment of unknown words, Cyclical Learning Rate to provide better convergence, Macro Soft-F1 Loss function to optimize performance and Macro Soft-F2 Loss function to deal with the imbalanced classification problem. RoBERTa-base model was also proposed, pre-trained on 58 million tweets and fine-tuned for ofensivelanguage identification. Experimental results show that, although it is a dificult task, both proposed approaches are suitable for detecting ofensive texts. Nevertheless, the Deep Learning approach achieves the best results.

Dissertação de mestrado em Crime, Diferença e Desigualdade
Com a evolução das novas tecnologias da informação e da comunicação, assistimos a um crescendo em atos de violência no contexto virtual. O cyberbullying é um fenómeno recente, de assédio e perseguição constante, com recurso às novas tecnologias, com destaque para a Internet. Através desta forma de violência virtual, as práticas ocorrem por meios tecnológicos e podem acontecer em qualquer momento, a qualquer hora e com qualquer pessoa. Regra geral, estes comportamentos verificam-se entre pares, contudo, afetam de maneira diferente crianças, jovens e adultos. No fundo, qualquer pessoa pode ser vítima, agressor ou testemunha desta forma renovada de bullying, cuja dimensão ainda não se conhece em concreto, mas que começa a ser frequente sobretudo nas redes sociais. Esta dissertação de mestrado tem como principal objetivo identificar, analisar e compreender as representações sociais dos estudantes da Universidade do Minho sobre o fenómeno do cyberbullying. Através de uma metodologia qualitativa foram realizadas um total de dez entrevistas semiestruturadas sobre as representações deste fenómeno. Para o efeito, segundo as conceções teóricas de Serge Moscovici procurou-se averiguar o nível de informação, as crenças, as atitudes e eventuais relatos de envolvência dos estudantes com o problema do cyberbullying. O presente estudo permitiu detetar alguns elementos muito próximos da posição de vítimas, agressores e testemunhas de episódios de cyberbullying. Segundo os participantes, as raparigas são mais identificadas como agressoras e vítimas deste fenómeno, por motivos que remetem para formas de violência mais subtis, indiretas e por uma excessiva exposição nas redes sociais. Os resultados encontrados com esta investigação revelam ainda outros pontos de vista marcados por opiniões contraditórias e até desconhecimento por parte dos estudantes relativamente a esta problemática. A nível geral, a investigação centrada na temática do cyberbullying deve permitir desenvolver abordagens práticas que garantam resultados positivos na prevenção e intervenção contra esta nova forma de vitimação.
With the evolution of the new technologies of information and communication, we witness a growing violence in the virtual context. Cyberbullying is a recent harassment and constant persecution phenomenon, using new technologies with an emphasis on the Internet. Through this form of virtual violence, acts occur via technological means and can happen at any moment, any time and with anyone. As a general rule, these behaviours occur between peers; however, they affect children, young people and adults differently. In the end, anyone can be a victim, aggressor or witness to this renewed form of bullying, whose dimension is not yet known, but which begins to be frequent mainly in social networks. This master's thesis has as main objective to identify, analyse and understand the social representations of the University of Minho students on the phenomenon of cyberbullying. Through a qualitative methodology, a total of ten semi-structured interviews were carried out on the social representations regarding this phenomenon. For this purpose, according to the theoretical conceptions of Serge Moscovici, we sought to ascertain the level of information, the beliefs, the attitudes and possible reports of students' involvement with the problem of cyberbullying. The present study allowed detecting some elements very close to the position of victims, aggressors and witnesses of episodes of cyberbullying. According to the participants, girls are more identified as aggressors and victims of this phenomenon, for reasons that refer to more subtle, indirect forms of violence and excessive exposure in social networks. The results found in this research also reveal other points of view marked by contradictory opinions and even lack of knowledge on the part of the students regarding this problem. In a broad sense, research focusing on the subject of cyberbullying should enable practical approaches to be developed that will ensure positive results in prevention and intervention against this new form of victimization.

The article examines the issue of improving the Communication Strategy of the Verkhovna Rada of Ukraine for 2017-2021 to increase the level of trust in Parliament, raise the political and legal culture of citizens, counteract information wars and operations against the Parliament, as well as the development of e-democracy. It is noted that the low level of public trust in Parliament causes the necessity of modernizing the parliamentary system of communications in the vector of approximation to EU standards and raising the political and legal culture of citizens. The thesis is grounded on the necessity of developing a communication system of protection against information wars and operations against the Parliament of Ukraine, forming its positive image and counteracting manipulative media. The main criterion for the effectiveness of the Strategy is proposed to be the increase of the level of trust in the Parliament and parliamentarians. It is suggested to pay more attention to the formation of the institutional basis for the implementation of the Communication Strategy: responsible entities, their job responsibilities and necessary resources. The implementation of joint projects of the Parliament of Ukraine with universities and research institutions, the establishment of a school of communications, conducting seminars, trainings and forums in cooperation with civil society will have a positive impact on the implementation of the Communication Strategy. This will develop the tradition of communication and understanding of the parliamentarian as a professional communicator. The implementation of the Communication Strategy envisages increasing the efficiency of the information space monitoring system, timely response to negative reports about the Parliament and prevention of their occurrence, which is the basis of anti-crisis communications. The tools that ensure the openness of the Parliament not only contribute to the realization of the idea of representation, but also immanently pose threats to democracy through manipulative media, trolling, cyberbullying. Digital communications are important, but by themselves will not completely solve the problem of low confidence in Parliament, and are seen as a modern tool that can be updated at any time. The implementation of the Communication Strategy of the Verkhovna Rada of Ukraine for 2017-2021 should be considered in the context of ensuring the national security of the state, branding of Ukraine in the world and responding to opportunities-threats from globalization.

In Japan, the use of mobile devices has been spreading rapidly and continues to grow. However, this causes many problems that no one predicted. This article describes the use of mobile devices as well as issues related to their use in Japan. The authors will discuss these issues regarding the three aspects of “sharing of information and emotions in emergent disasters,” “addictive use of mobile devices,” and “cyberbullying in young people,” and will introduce representative psychological researches in Japan. At the end of this article, the actions and directions necessary to address these issues will be discussed.

The phenomenon of stalking itself is as old as humankind itself. The past years and the expansion of digitalization have put the crime of stalking in a new light. A representative study from the USA shows that one in four Americans has already been “harassed” online. If stalking in the traditional sense (also known as offline stalking or classic stalking) is described as obsessive harassment/threat over a longer period, it takes on a new criminal dimension as soon as connected communication takes place. New forms of crime such as cyberstalking, cyberbullying, and romance scams have now reached our everyday lives. This article explores cyberstalking.

The article attempts to consider the phenomenon of social necrophilia, described by Erich Fromm, from the angle of transformations taking place in the world in the digital age. Digitalization intensifies necrophilic tendencies of modern civilization: mechanization and robotization, dominance of the material, fixation on the past, etc.; contributes to an outbreak of interest in the once taboo subject of death, so it’s worth talking about digital necrophilia. It is noted that the virtual mode of being creates a wide range of opportunities for a fixed representation of one's Self (social networks), for the release of aggression (cyberbullying, cancelling), for the reorganization of funeral culture and the process of mourning (digital cemeteries, memorials). The anthropologization of technology, the formality and coldness of interpersonal relationships, transgressive games and cinema, the translation of long speech (Dialogue in its hermeneutic presentation) into a signed (fixed) form - all

The empirical study involves detecting vulnerable images in social media. Use a vulnerable image detection dataset with a 78% data accuracy rate using a graph neural network. The study investigates that sharing vulnerable pictures on social networks might cause cyberbullying or stalking victims leads to many stressful events. The inexcusable behavior by many people is encompassing in the digital world. This phenomenon leads to threats of violence, slander, or even leakage of personal images may be included. The study associates vulnerability detection with the pictures posted on social media. In this way, the people of the society will be able to identify the vulnerable images and help to build social awareness. The ability of GNN to model the dependence between nodes in a network allows for a breakthrough in graph analysis research. A graph autoencoder framework can help GNNs solve the network embedding challenge. The approach associate's construction of a custom dataset with two classes, involve Vulnerable and Nonvulnerable images. We declare image size and batch size. After creating the GNN model data is trained using 10 epochs and then plot the accuracy and loss curve. The significance of the accuracy rate reached a satisfactory level. For the statistical approach, a confusion matrix is constructed to compare the actual and prediction data this helps to understand the performance of the proposed model. Graph neural network is very effective to classify images and other things. The findings constructed a Graph Neural Networks (GNNs) is an effective framework for representation learning of graphs.

Communication using modern internet technologies has revolutionized the ways humans exchange information.. Despite the numerous advantages offered by such technology, its applicability is still limited due to problems stemming from personal attacks and pseudoattacks. On social media platforms, these toxic contents may take the form of texts (e.g., online chats, emails), speech, and even images and movie clips. Because the cyberbullying of an individual via the use of such toxic digital content may have severe consequences, it is essential to design and implement, among others, various techniques to automatically detect, using machine learning approaches, cyberbullying on social media. It is important to use word embedding techniques to represent words for text analysis, typically in the form of a real-valued vector that encodes the meaning of words. The extracted embeddings are used to decide if a digital input contains cyberbullying contents. Supplying strong word representations to classification methods is a key facet of such detection approaches. In this paper, we evaluate the ELMo word embedding against three other word embeddings, namely, TF-IDF, Word2Vec, and BERT, using three basic machine learning models and four deep learning models. The results show that the ELMo word embeddings have the best results when combined with neural network-based machine learning models.