Abstract

In the contemporary digital landscape, cyberbullying continues to pose a significant threat to online safety and well-being. This paper discusses the technological advancements and methodologies behind the project and presents compelling evidence of its efficacy in transforming the landscape of cyberbullying recovery. To ad-dress this pressing issue, we introduce the "Cyberbullying Advisory Chatbot," a pioneering AI-powered solution. Unlike conventional technologies, our chatbot is uniquely designed to provide a dedicated platform for cyberbully-ing victims from various online platforms, offering them a space to share their experiences and embark on a jour-ney of emotional recovery. Notably, it extends its support by providing legal guidance to help users navigate the complexities of online harassment. Including Natural Language Processing (NLP), Natural Language Understand-ing (NLU) and Deep Learning. Making use of task-specific datasets, these models employ the key components such as Toxicity Level Detection, Empathetic Response, Intent Recognition. Leveraging BERT Model for enhanced user interaction in chatbot interface. These technologies empower the chatbot to offer real-time empathetic re-sponses and tailored legal advice, thereby redefining the user experience in confronting cyberbullying. The integra-tion of AIdriven analysis, coupled with supervised learning techniques, ensures the recognition and categorization of toxic content with exceptional accuracy, enabling users to receive personalized support. The novelty of our pro-ject lies in the harmonious synergy of these components, prioritizing accuracy through the usage of task-specific datasets. By empowering victims with empathetic and context-aware responses, our chatbot acts as a resilient ally, offering support to those affected by cyberbullying across various digital platforms.