Big Data

Election Tech: Can big data and social media predict a contested convention?

Social media data analysis shows a recent, rapid spike in voter chatter about the likelihood of a GOP convention smackdown.

contestedconventionconversation.png
Contested convention conversation trend. | Image: Sysomos

Online conversation about a GOP convention fight has gone mainstream. Discussion about a contested convention has been trending since late February, and as polls closed Tuesday night convention fight conversation spiked on blogs, news sites, and social media platforms, said big data firm Sysomos.

The company started indexing campaign-related content in early March. The social media tool performs Boolean searches and hunts for terms related to candidates, elections, and issues. The contested convention trend emerged early, as polls were closing on Tuesday.

Sysomos indexes news sites and blogs, as well as Twitter, Facebook, Instagram, YouTube, and other large social media services. The platform also finds relationships between various types of web content and allows users to perform complex searches against its database. The database itself includes over 500 billion unique conversations and uses algorithms to "measure sentiment, identify key influencers, measure geo-location, identify trends, and surface popular content." The raw data is articulated using various content-specific dashboards.

READ: The Power of IoT and Big Data (Tech Pro Research story)

Between March 13th and March 15th there were nearly 530,000 unique mentions of the mid-month primary election. Approximately 9.8% of those conversations were about a "Contested Convention" OR "Brokered Convention" OR "Primary." Interesting, according to the database, 86% of the mentions were by men, and only 14% by women.

contestedconventionwords.png
These are the words most closely related to the phrase 'Contested Convention' on social media. | Image: Sysomos

The trend began about three weeks prior to the March 15th primary. On February 25th, the day of the CNN Republican debate, there were 5,767 online mentions of a contested convention. The number of mentions grew to 21,803 on March 4th, when Ben Carson dropped out of the race. By March 15th, primary day, conversation jumped to 52,406 mentions.

sysomoselectiontrend.png
Related election trends. | Image: Sysomos

Donald Trump is a global candidate and is discussed around the world like no other candidate. Trump was mentioned 4.3 million times in the days preceding the March 15 primary. Trump's gender trends are also fairly evenly distributed between men and women. 58% of Trump chatter was by men and 42% by women.

trumpglobalmentions.png
Distribution of traditional and social media mentions of Donald Trump from March 13th - March 16th, 2016. | Image: Sysomos

Ted Cruz was a distant second, with 683,247 mentions, 58% by men and 42% by women. Bernie Sanders earned 640,968 mentions and was discussed 59% by men and 41% by women. Kasic and Rubio each had fewer than a half million mentions.

Over the course of the campaign we will continue to perform simple data analysis. We hope to uncover unique insights and find ways business can benefit from campaign innovations.

If you're a data scientist, social media professional, or inquisitive TechRepublic reader we'd love your ideas on how to inspect campaign social media data. Please leave a comment below or ping us on Twitter @TechRepublic.

Read more

About

Dan is a Senior Writer for TechRepublic. He covers cybersecurity and the intersection of technology, politics and government.

Editor's Picks