FIRST AIRED: June 3, 2016

Nice work! Enjoy the show!

×

You’re busy. We get it.

Stay on top of the news with our Editor’s Picks newsletter.

US Edition
Intl. Edition
Unsubscribe at any time. One click, it’s gone.

Thanks for signing up!

We've got more news

Get our editor’s daily email summary of what’s going on in the world.

US Edition
Intl. Edition
Replay Program
0:00
0:00
More Info

COMING UP:Share Opener Variant 1

×

Transcript

00:00:00
for years politicians have polled voters and constituents with phone calls and focus groups , now they're turning online to social media to measure zeitgeist but the algorithms they used to mind that data come with a few hiccups , terry mcbride is covering the story for writers won't problem is they just don't understand sarcasm and there's a lot of sarcasm , if you and i were talking and one of us said , yeah right or short as you know immediately and contacts that that list meant sarcastically but an algorithm doesn't know that so it would probably take you seriously so now data scientists who work for political consulting firms are training algorithms to recognize certain combinations of words as negative , the word classy alone positive , a pair that was one of the branches figures to enter the twenty sixteen presidential race not so much , aden stay classy and trump and a hundred percent at that , it's negative campaigns reliance such research to help them tailor advertising and other outreach the particular groups of voters are crucial tool in the modern american political camp , sarcasm clues can also come from the identity of the poster , for example a seemingly pro republican tweeter facebook posts coming from someone who follows mostly democrats , the trigger and alerting the algorithm that the evaluation of the tweet might not be accurate