Classifying Social Media Comments Using Machine Learning

Abstract: The demand to reach and satisfy audiences world- wide increases the number of influencers and content creators on social media, which is the primary platform to disseminate their work. Each video could potentially get thousands of comments as a content creator grows, and these comments acts as direct feedback from the viewers, also as major means of understanding viewer expectations and improving channel engagement. We have proposed approach to classify social media comments into five categories namely good, discussion, motivational, demotivating and abusive. In this paper we have elaborated comparative analysis between the available machine learning classification algorithm like Logistic Regression, SGD Classifier, and Random Forest. Index Terms: Machine Learning, Natural language Process- ing, Logistic Regression, SGD Classifier, Random Forest