In this study, we attempt to identify the emotion levels, such as positive, negative, & neutral feelings, from postings and comments on social networking sites on depression. Social media sites like Facebook and Twitter are becoming effective for helping those in need who require extra care or attention in terms of mental support. They are also utilized for communication and network development among relationships. There are several depressive support groups on Facebook, and they are quite helpful in giving the sufferers mental assistance. In this study, we attempt to formalize the posts and comments on depression into a succinct lexical database and identify the emotion levels from each occurrence. The complete amount of work has been divided into two sections: sentiment analysis and the use of machine learning techniques to examine the capability of extracting sentiment from such a unique category of texts. To determine the sentiment levels, we used the Python textblob module and typical machine learning techniques on the linguistic characteristics. For each of the classifiers, we have calculated the precision, recall, F-measure, accuracy, and ROC values. Random Forest outperformed the other classifiers, successfully classifying 60.54% of the instances. We think that conducting sentiment analysis on a particular class of texts may inspire additional research into how natural language is understood.