The exponential increase in the use of social media jargons and trend of fast paced conversations has led to an increase in the practice of slangs and vulgar language among netizens, for various different contexts. These use of vulgar words and over sexualization of expression in conversation causes a loss of depth in them, which also negatively affect our cognitive conditioning and Linguistic Intelligence, and majorly affect children as they lack the Emotional, Cognitive and Integral Intelligence, which is basically required to differentiate between right and wrong. Machine learning can help us detect vulgar language in texts on digital platforms, for further understanding and study with the intention of creating a healthier cyber world in the realm of digital communication and expression. This research work focuses on detecting vulgar language in texts, on social media. The study has evaluated four different machine learning classifiers by training them on the same web scraped X (formerly Twitter) dataset using supervised machine learning approach. The evaluation of the proposed approach on X dataset shows that Logistic Regression and Naïve Bayes achieves 94% accuracy, whereas Decision Classifier and Random Forest achieves 96% accuracy, on the same test dataset.