Overview paper on social media analytic in predicting elections
Mohamed Tabouti
MIS 2501
Extra credit assignment (Social Media Analytic to Predict Elections)
One of the latest innovative ways to leverage the power of social media is mining the data from the most popular social media platforms to predict elections. The field is in its infancy stage and while optimistic stakeholders are working hard to polish it. Skeptics have multiple arguments on why it’s not worth it to invest research efforts in that field.
The way the process works is through putting together sophisticated algorithms that screen popular blogs, sites, Facebook pages and Twitter feeds to gather buzz words, trends, likes, shares, and even text data from discussions, tweets and posts. A correlation has been shown to exist between politicians who generate the most buzz on social media and those who actually win the elections. A recent example was on 2012 presidential elections, where the predictions from social media analytic successfully predicted President Barack Obama as a winner. And so it was. This could be the start of a revolution on how pooling and sentiments are done. It also proves the importance of social media as a channel to create buzzes and influence public opinions.
While the idea is still riding the inflated first stage of the hype cycle. Skeptics are pointing out reasonably formulated arguments on why social media shouldn’t be a reliable source of predictive data. The idea is that no algorithm, at least up to date, will be able to detect the difference between a sincere comment and a sarcastic one or a joke. Individuals maybe commenting sarcastically “Yeah right, I will vote for Romney” and the algorithm would naturally interpret that as a favorable comment for Romney, while it was a sarcastic way to express the inverse. Also, various demographics are not sufficiently represented in social media which could potentially skew the findings. Going back the last presidential elections, Ron Paul has generated a big buzz in social media because it was a big focus to his campaign, however, when come elections day. His ranking in actual votes didn’t match his predicted popularity.