POTUS.
Mr. President.
Tweeter-in-Chief.
Depending on who you ask, Donald Trump’s twitter is a constant source of … something. Be it amusement, disbelief, grief, agony, comedy. I’m never sure if I should laugh or cry or both, and I know I’m not the only one.
Thanks to the team at Synthetic, a dedicated cognitive technology practice and a division of Omnicom-owned digital marketing agency, Organic Inc., we now have some musical insight into what’s behind the 220 characters.
“We were talking about Trump and how nobody knows what’s going on in his mind,” said Keith Pine, Chief Operating Officer, Organic Inc. “We wanted to use AI to get inside his head and figure out the emotions behind his tweets.”
As a vehicle, music elevates language to another dimension, and through sentiment analysis and natural language processing, Synthetic created an AI that extracts attitudes and emotions from Trump’s tweets that only a machine can detect.
That sentiment is then paired with songs with similar sentiment using the same technological process – so when The POTUS tweets, Synthetic processes the sentiment and posts a corresponding song through the @TrumpMasterFlex Twitter account. At the end of the week all songs are aggregated in the Trump Master Flex playlist on Spotify.
“Basically, we are using sentiment analysis and natural language processing to understand what he’s saying and to find commonalities in theme, meaning, and emotions,” Pine said. “We then run the same process against a giant lyric database to find songs that match the sentiments of the tweet.”
? ? ?
After extensive natural language processing, I’ve determined Bills, Bills, Bills best matches this tweet https://t.co/atfS5yYWC0
— TrumpMaster Flex (@TrumpMasterFlex) March 23, 2018
It’s almost not fair how funny @TrumpMasterFlex is, but on a larger scale, this kind of applied AI technology can help us determine sentiments far beyond Trump’s tweets. Consumers talk about brands and products online all the time, and they do so publicly. Being able to extract extensive amounts of sentiment data across a variety of social platforms can mean big things for marketers, and Pine agrees.
“We just thought this would be a fun idea to get out into the world, but now we’re working on a couple of tools that use a similar methodology to understand who our customers are,” said Pine.
And it all came from wanting to understand why our President tweets from the West Wing at 3 am.
2k18, am I right?
