Do the political opinions shared on Social Media (platform Twitter) accurately represent the political opinions of the general populace? (original) (raw)

This research investigates the relationship between political opinions expressed on Social Media and the political opinions of the public (electorate) on major political event. Specifically, on the 2016 UK EU Referendum vote. It looks to answer the question, are Social Media sites such as Twitter echo chambers of people with the same opinion or do they accurately represent the opinions of the public. The investigation includes a computer program that extracts a Big Data set of tweets from Twitter, then analyses them to find their binary sentiment using a democracy of multiple machine learning sentiment classifiers. This is then compared to a database of pollster data from a variety of pollsters to find any differences. It also investigates why there is or isn’t a difference in opinions between Twitter users and the public.