The proliferation of spam content is on the rise due to the widespread use of social media. Users receive numerous text messages via social media platforms, making it challenging to identify spam within these messages. Spam messages often include harmful links, deceptive apps, fraudulent accounts, fake news, misleading reviews, and rumors. In this sense, enhancing social media security necessitates the crucial task of detecting and controlling spam text. This paper presents a practical approach for classification of spam detection in social media comments using python programming. Using large datasets from Facebook, Twitter, YouTube, SMS, Reddit and E-mail and using the Decision Trees, Logistic Regression and Random Forest it was achieved accuracy between 88% and 96% of spam classification. The datasets and implementations are available on the opensource platform Ghitub to be used and improved in future works.