2018 Pakistan General Election: Understanding the Predictive Power of Social Media
Maryam Bilal, Sundas Asif, Shainila Yousuf, Uzma Afzal · 2018
Social media has become ubiquitous and globally used for social networking and content sharing. Twitter is a well-known micro blogging social media site to share and exchange people's opinion and content in different aspects of life. The power of twitter based predictions has its own magnified importance and has been greatly accepted. In this research paper, we investigate a tweet-based application of deep neural network to predict the 2018 electoral results of Pakistan. We collected tweets data using a twitter API. First, we tuned the hyper parameters of neural network then we trained the model with the tweets dataset of 2013 elections. We tested the trained model with tweets data set of 2018 election, model showed PTI as a clear winner. Results are very close to the actual outcome of the 2018 elections which also declared PTI as the biggest political party.