Analysing and Predicting Propaganda on Social Media using Machine Learning Techniques

Social Networks has overwhelmed the entire world by giving a foundation of data dispersal. People in groups share data without knowing its authenticity. These days Social Networks are utilized for picking up impact in numerous fields like in elections, advertisements and so on. It isn't amazing that Social media has become a weapon for controlling sentiments by spreading disinformation. Some clients share data intentionally in order to impact the crowd. Propaganda is one of the Systematic and purposeful endeavors utilized for impacting individuals for the political and religious gains. Data is collected from online news sources and from Twitter (Social Networking Site) using Twitter API. In the wake of explaining the content, methods like Term Frequency/Inverse Document Frequency (TF/IDF) and Bag of words (BOW) are used. Machine Learning classifiers are used for classifying the text into binary classes (Propaganda and Non Propaganda). Support vector machine showed the better results among all other traditional machine learning algorithms.