This paper investigates the social networks usage by students in Abidjan city, Côte d’Ivoire. We focus on a descriptive or quantitative analysis to understand the motivations and how students make use of in...This paper investigates the social networks usage by students in Abidjan city, Côte d’Ivoire. We focus on a descriptive or quantitative analysis to understand the motivations and how students make use of internet and social networks. More than six hundred forms were distributed to persons we have deemed as students. In return, we received more than 93% of the forms that have been processed. The study highlights the materials and the digital platforms that students used the most. The majority of the respondents reported to have access to the social networks in their mobile phones, with WhatsApp leading this application ranking, followed by Instagram, Facebook, YouTube, and Tik Tok. The survey shows that two third of our respondents are aged from 19 to 25 years old and almost half of the respondents spend daily 2 to 5 hours on digital platforms. The investigation also reveals that the main online activities are the e-commerce, chatting, information, and entertainment. The paper addresses also the online harassment of the students and it shows that more than one tenth of them have been victims of cyber-bullying. This study might be useful for governments, institutions, academia, individuals and professionals in order to communicate efficiently with a given population for a better use of social networks and to prevent students from harassment.展开更多
随着社交网络平台的迅速发展,网络欺凌问题日益突出,文本与图片相结合的多样化网络表达形式提高了网络欺凌的检测和治理难度.构建了一个包含文本和图片的中文多模态网络欺凌数据集,将BERT(bidirectional encoder representations from t...随着社交网络平台的迅速发展,网络欺凌问题日益突出,文本与图片相结合的多样化网络表达形式提高了网络欺凌的检测和治理难度.构建了一个包含文本和图片的中文多模态网络欺凌数据集,将BERT(bidirectional encoder representations from transformers)模型与ResNet50模型相结合,分别提取文本和图片的单模态特征,并进行决策层融合,对融合后的特征进行检测,实现了对网络欺凌与非网络欺凌2个类别的文本和图片的准确识别.实验结果表明,提出的多模态网络欺凌检测模型能够有效识别出包含文本与图片的具有网络欺凌性质的社交网络帖子或者评论,提高了多模态形式网络欺凌检测的实用性、准确性和效率,为社交网络平台的网络欺凌检测和治理提供了一种新的思路和方法,有助于构建更加健康、文明的网络环境.展开更多
文摘This paper investigates the social networks usage by students in Abidjan city, Côte d’Ivoire. We focus on a descriptive or quantitative analysis to understand the motivations and how students make use of internet and social networks. More than six hundred forms were distributed to persons we have deemed as students. In return, we received more than 93% of the forms that have been processed. The study highlights the materials and the digital platforms that students used the most. The majority of the respondents reported to have access to the social networks in their mobile phones, with WhatsApp leading this application ranking, followed by Instagram, Facebook, YouTube, and Tik Tok. The survey shows that two third of our respondents are aged from 19 to 25 years old and almost half of the respondents spend daily 2 to 5 hours on digital platforms. The investigation also reveals that the main online activities are the e-commerce, chatting, information, and entertainment. The paper addresses also the online harassment of the students and it shows that more than one tenth of them have been victims of cyber-bullying. This study might be useful for governments, institutions, academia, individuals and professionals in order to communicate efficiently with a given population for a better use of social networks and to prevent students from harassment.
文摘随着社交网络平台的迅速发展,网络欺凌问题日益突出,文本与图片相结合的多样化网络表达形式提高了网络欺凌的检测和治理难度.构建了一个包含文本和图片的中文多模态网络欺凌数据集,将BERT(bidirectional encoder representations from transformers)模型与ResNet50模型相结合,分别提取文本和图片的单模态特征,并进行决策层融合,对融合后的特征进行检测,实现了对网络欺凌与非网络欺凌2个类别的文本和图片的准确识别.实验结果表明,提出的多模态网络欺凌检测模型能够有效识别出包含文本与图片的具有网络欺凌性质的社交网络帖子或者评论,提高了多模态形式网络欺凌检测的实用性、准确性和效率,为社交网络平台的网络欺凌检测和治理提供了一种新的思路和方法,有助于构建更加健康、文明的网络环境.
