Download PDFOpen PDF in browserCyber Bullying Detection on Social MediaEasyChair Preprint 100084 pages•Date: May 9, 2023AbstractNow a days peoples use social media to create, share and exchanges information and ideas in virtual communities and network.For examples Instagram, Facebook, Twitter, etc...As the technology grows the negative part CYBER BULLYING is also grown with it.It's nothing but collecting of personal information of users and emotionaly misuse it in the way of online harrasment, defame any person and online financial fraudness with stranger profile. To prevent from this kind of negative we build a Machine Learning approach system model to detect cyberbulling by analyzing the emotional content of text.By using datasets of online conversations to train and test a model that classifies text as either cyber bullying or non cyberbullying.The model uses Natural Language Preprocessing techniques such as sentiment analysis and topic modeling to identify patterns of abusive language and offensive content.The model accuracy is evaluvated using precision recall and is formed to be effective in detecting cyberbullying with an accuracy of 87%.The rescus suggest that emotion analysis can be an effective tool for detect cyber bullying and may help identify and prevent harmful behaviour online. Keyphrases: Cyber bullying, Natural Language Processing, machine learning
|