Now that the election has come and gone, it’s time for reflection. Donald Trump’s surprising victory over Hillary Clinton was a shock to the entire world. Fortune got it right, saying “humans failed, not the data.” What’s clear is that the establishment media and political pundits got it wrong, neglecting the true metrics that were showing a groundswell of support for the Republican ticket. The pollsters looked past what the data was telling them, and though Donald Trump was the eventual winner, Big Data was triumphant. What needs work is the Data Analytics and predictive forecasting methods that were employed by many of the major news outlets.
Predata, a firm that condenses data sources from around the web into clear and unified signals for geopolitical risk found that Donald Trump was well ahead in the digital conversation in the 12 major battleground states during the final week leading up the election, winning by a digital campaign score of 89.3 percent to 67.8 percent (Clinton). Predata worked on the presumption that political campaigning increasingly takes place online, and developed signals to capture shifts in conversation around the race. They gathered hundreds of thousands of data points on a daily basis. Although the firm got their prediction wrong, the company admits that they ignored many of the signs that pointed toward a Trump victory, citing human error, adding “the data was blameless.”
Similarly, an artificial intelligence system called MogIA developed by Sanjiv Rai, founder of an Indian startup Genic.ai, correctly predicted a Donald Trump victory on November 8th. The AI system has been correct in predicting the last 3 elections starting in 2004. MogIA takes into account 20 million data points from public social media platforms including Google, Facebook, Twitter and YouTube and runs analysis on the data to create predictions. Not only did Trump dominate in terms of online queries in the race against Hillary Clinton, Trump shattered the highs recorded by Barack Obama’s 2008 campaign by more than 25 percent. According to Rai, his AI technology shows that the candidate who leads in social media data has been a perfect predictor of who will win. Judging by the numbers the AI recorded in 2016, Donald Trump generated the most buzz of any candidate ever recorded.
Barack Obama successfully utilized Big Data analytics in both 2008 and 2012. His campaign used these tools to deliver data forecasting and predictive modeling to help drive decision-making and was certainly a driving factor in his two electoral victories. The Washington Post even labeled Obama as the “Big Data President” for his campaign’s role in utilizing these resources on a mass scale. The campaign employed more than 100 data analysts who ran upwards of 66,000 computer simulations on a daily basis, with Obama’s adviser Jim Messina wanting to “measure everything.” Similarly, Hillary Clinton’s campaign also utilized artificial intelligence to make decisions about how the Democratic ticket should function in this year’s contest.
No matter your political lean, the era of Big Data dominating politics is here. What remains to be seen is whether or not those that report on the findings can avoid the human bias that was so prevalent this time around.
Latest posts by Timothy King (see all)
- Informatica Merges Data Management and AI in Platform Update - September 14, 2017
- 3 Data Lake Deployment Best Practices - September 8, 2017
- Dataiku Nabs New Funding for Data Science Democratization - September 6, 2017