Social Network based Sentiment and Network Analysis to Predict Elections

Abstract

The General Elections in India is a grand event with over 900 million eligible voters. India has a population of around 1.3 billion and it is the world’s largest democracy. One of the features of a democratic nation is the election. In election general sentiment and opinion of people plays an important role. Social media is abuzz with opinions and discussions taking place all over India especially twitter. We tried to predict election results and general mindset using sentiment and network analysis. Previously, some work has been done in the sphere of gauging the mindset of people using social media, specifically Sentiment Analysis of Twitter posts. Using sentiment analysis, we can classify a document into a host of categories like positive, negative, neutral, angry, happy, sad, etc. and this is especially helpful when we are trying to gauge the opinion of people based on what they wrote about it. We used Twitter data for our work and tried to apply sentiment analysis techniques to find the consensus of the public about different political leaders and political parties. The results were pretty accurate when compared to actual results and this model and analysis could prove useful in gauging the mindset of the general public about a specific topic, product, or idea. Furthermore, we also carry out a detailed network analysis of dataset using the network formed by taking tweets as the edges of the graph. This gave us some really interesting insights and resulted in the formation of 4 major communities confirming the fact that online communities are close-knitted when it comes to political orientations.