In this information era, social media has become a most popular means of communication and the source of news. Its usage is growing rapidly and as a result, varieties of contents are shared on these platforms. As social media has grown closer to everyone’s everyday digital lives, it has become a platform where people publish their ideas or opinions directly, sometimes going beyond the bounds of communication standards such as bullying. Currently the most widely used method to detect text-based cyberbullying is the use of machine learning models, especially the text-based sentiment analysis models. As the research area gained more attention, wide range of machine learning and deep learning models are introduced for sentiment analysis. So it is important to also understand the effective approach among the most widely used models and the improvements that are needed to be made in the approaches. In this work, an effort is made to identify the sentiment behind the textual social media content and classify them depending on their polarity using some well-known traditional and deep learning models. Thereby demonstrating the application of machine learning-based model in identifying any attempt of cyberbullying in social media platforms. The work also presents how machine learning models work on par with deep learning models.