Election Tweets Prediction Using Enhanced Cart and Random Forest

Political system of a country has always been complicated in nature, and this complexity can be due to various factors such as number of parties, policies and, most notably, mixed public opinion. The advent of social media has given people around the world the ability to converse and discusses with a very wide audience; the sheer amount of attention gained from a tweet or a post is unimaginable. Recent advances in the area of profound learning have contributed to their use of many different verticals. Techniques such as long-term memory (LSTM) allow a sentiment analysis of the posts to be carried out. This can be used to determine the masses’ overall feelings towards a political party or person. Several experiments have shown how to forecast public sentiment loosely by examining consumer behaviour in blogging sites and online social networks, such as in national elections. Machine learning has a rapid growth in recent years and has been applied from self-driving cars to e health sectors in every technology. A model of machine learning is proposed to predict the chances of winning the upcoming election based on consumer or supporter views on the web of social media. The supporter or user share their opinion or suggestions for the group or opposite group of their choice in social media. The text posts are needed to be collected about election and political campaigns; the models of machine learning are developed to predict the outcome.