Abstract: Due to the expansion of the Internet, the usage of web-based enjoyment has grown dramatically over time and is currently the dominant organising platform of the twenty-first century, surpassing conventional media. However, the increased social accessibility frequently has bad repercussions on society that combine two or three awful traits, such as online abuse, inciting cyberbullying, cybercrime, and web-based savaging. Cyberbullying frequently has negative consequences, particularly for girls and kids. in severe emotional suffering, which may still prompt suicidal thoughts. Online haranguing stands out because of the profound negative impact it has on society. Online harassment has recently resulted in a number of things, including the dissemination of sexual comments, rumours, and private chats. Analysts are thus paying closer attention to the identification of abusive SMS or messages from video streaming.
The goal of this project is to combine natural language processing and machine intelligence to design and implement a practical strategy for identifying harmful and harassing online postings. Bag-of-Words (Bow) and word recurrence reverse text recurrence are two unique characteristics that are used to assess the correctness of four different AI systems (TFIDF).

Keywords: Cyber-Harassing, Natural Language Processing, Machine Learning and social media.

| DOI: 10.17148/IARJSET.2022.9673