Analyzing Information Flow of Hashtag Networks during Elections using Sentiment Analysis and Graph Algorithms

An exponential increase in the usage of social media across the world creates a lot of unstructured data and cross-communication between individuals. These platforms provides opportunity to the political parties to spread their word out. The information is spread using several hashtags in the form of user-generated tagging that facilitates cross-referencing of content. These hashtag-generated networks serve as a huge reservoir of data and if analyzed systematically can help in understanding the agenda-setting of each party and how successful or unsuccessful they are. This in turn helps in predicting the outcome of the election looking from the prism of social media. In the present study, a model is proposed by combining sentiment analysis and graph techniques to look into the trending hashtag networks propagated by political parties using Twitter. The sentiment analysis gives us a sense of inclination of each tweet and thereafter it’s extrapolated onto the hashtag’s user network to get insights as to how the information is diffusing and how one party propagates its favorable hashtag and how the others try to counter it. The major aim of the present work is to find out the intricacies that go on in the social media space before a major election.