Social Movement Prediction from Bangla Social Media Data Using Gated Recurrent Unit Neural Network

Social networking platforms have become an essential part of our everyday life. These sites contain more data than any other platform now. Analyzing these social interactions can prove to be a key factor in predicting social movements. In this research work, we used the Gated Recurrent Unit (GRU) Neural Network for analyzing Bangla social media data to predict social movements. We prepared a dataset that represents an equal number of negative and positive notions about social movements. We gathered data from Twitter, which is currently one of the most renowned social networking platforms around the globe. We showed how hyper-parameters optimization can be profitable to attain the best probable outcome from GRU-RNN. This research work will help to figure out whether any social movement is going to take place or not and also help the government or the law enforcement agency to fulfill the public demands or take necessary steps long before the social movements take place.