Analysis of Political Sentiment Orientations on Twitter

The dramatic increase in the number of users on social media platform leads to the generation of huge amount of unstructured text in the form of messages, chats, posts and blogs. Besides the exchange of information, social media is a remarkably convenient medium to express the ideas and opinions which gain popularity when liked by a large set of users. This popularity may reflect the sentiment of people towards that person, organization or a place. The social media platform, such as Twitter, generates huge amounts of the text containing political insights, which can be mined to analyze the people’s opinion and predict the future trends in the elections. In this work, an attempt is made to the mine tweets, capture the political sentiments from it and model it as a supervised learning problem. The extraction of tweets pertaining to the General Elections of India in 2019 is carried out along with the study of sentiments among Twitter users towards the major national political parties participating in the electoral process. Subsequently, the classification model based on sentiments is prepared to predict the inclination of tweets to infer the results of the elections. The Long Short Term Memory (LSTM) is employed to prepare the classification model and compare it with the classical machine learning models.