Cyberbullying detection in Hinglish comments from social media using machine learning techniques

With the development of the Internet, the use of social media has increased dramatically over time and has emerged as the most powerful networking tool of the twenty-first century. From youngsters of ten years to senior citizens of sixty years, everyone is profoundly active in social media. Social media, due to its easy accessibility, has become a major part of our lives in all segments. Social media became a crucial platform for communication during the COVID-19 pandemic as people were socially isolated and had little access to others. The use of social media platforms helped keep the world connected. Those who were stuck at home alone turned to social media to keep in touch with friends and find entertainment. However, with the positives of social media come along a very dark negative side too. The greater involvement of social media has given rise to cyberbully where someone bullies or harasses others over social media. People, especially teenagers are found to write and post negative comments in facebook, instagram, youtube etc. This has become a major social cause as such activities are extremely disturbing for the victim. In this paper, we have proposed a comparative study between various machine learning frameworks used to detect cyberbullying in social media comments. In addition, the comments are also classified according to the severity of cyberbullying. The main contribution of this work is to collect a sizable data or comments based on the Hinglish language (a mixture of English and Hindi terms written in Latin script) and then detect cyberbullying in the Hinglish language. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.