Your Computer Understands Sarcasm? Yeah, right.

Detecting sarcasm is an essential skill in understanding the people around you. So much of what we see on the internet is sarcastic and joking. Humans can instinctively pick up on saracasm in a social media post and make a mental note that it isn’t serious. But how do serious comments get sorted from sarcastic ones by something that doesn’t understand its nuances, like a computer?

Think about Sheldon Cooper in the Big Bang Theory. He is a brilliant astrophysicist but he can’t understand simple social cues. A running joke in the show is that he often doesn’t understand what’s going on around him because people are being sarcastic, he takes them too literally, and then has to be told the comment was sarcastic so he can react appropriately. Sheldon Cooper acts much like a computer does. It is brilliant at understanding literal information but doesn’t understand a joke.

For class a couple of weeks ago we were assigned to watch Del Harvey’s about protecting Twitter users from themselves and others. She mentioned an example about a tweet with the text “yo bitch” that had a picture of a dog attached. Without the picture this tweet is fairly aggressive if you don’t understand culturally relevant dialogue and references. But with the picture Del Harvey understands it is a joke. What she doesn’t mention, however, is who or what is analyzing this single tweet. Harvey can’t personally look over every flagged tweet because as she mentioned, there are over 500 million tweets a day.

What social media scanning is already happening?

Facebook

Facebook monitors user chats and posts for predatory or illegal behaviors. It has turned in many possible suspects to law enforcement officials. One predator was a 30 year old man inboxing a 13 year old in South Florida about sex and planned to pick her up from middle school the following day. The conversation was automatically flagged and Facebook employees looked at it and contacted authorities to arrest the man. Facebook uses software to scan conversations for certain criteria and behaviors.

  1. The users aren’t friends, only recently became friends, or have no mutual friends
  2. They interact with each other very little
  3. They have a significant age difference
  4. They are located far from each other

Along with this criteria they look for phrases used by past criminals. A Facebook employee will not actually look at a post or message unless some of these things occur.

The Super Bowl

The Department of Homeland Security classified the 2015 Super Bowl in Phoenix, AZ as a Level 1 security event, only behind events like the U.N. General Assembly. With over 60,000 people traveling to the event and the general threat of bombings and shootings in large gatherings that plagues our world today, The Secret Service reported that they would be monitoring Twitter and Facebook for threats. At the time, a Service Service spokesperson told the Washington Post that screening for sarcasm was “just one of 16 or 18 things we are looking at.” But, they were unable to acquire software that would  help detect for sarcasm and pick out fake threats. Strategies like this were also used during the Boston Marathon following 2013

And so, a sarcastic machine was developed!

Rossano Schifanella, a computer science professor at University of Turin, along with colleagues from Yahoo!, worked to teach computers about sarcasm. The team did a research study involving English speaking participants where they had to mark social media posts as sarcastic or not. First, just text posts and then text that also had images.What they found was that images, linguistic cues and word play (“I just loooooove snow” versus “I just love snow”), and punctuation (especially exclamation points) showed sarcasm the most. They then created a mathematical algorithm based on what they learned from humans that detected sarcasm accurately 80 to 89 percent of the time. That’s pretty good for a machine.

What sort of opportunities does this present ?

Think about marketers that reveal a new product. Often, they turn to data analytics of social media posts to understand its success. How many times is the product mentioned positively? How many times is it mentioned negatively? The numbers could be skewed by sarcasm.

I think the biggest and most widespread sarcastic campaign I’ve seen online is the #ThanksObama. If you’re wondering what it is Urban Dictonary can tell you:

Screen Shot 2016-10-20 at 1.12.09 AM.png

There must be hundreds of tweets a day that mention this hashtag and not all of them can be counted in as praise. Of course we humans know this but a normal computer wouldn’t. A simple analytical tool might say based on Twitter activity that Obama has a high approval rating.

Rossano Schifanella’s algorithm could help reduce wasted time and provide more accurate data analytics. The content on the internet is only increasing exponentially so more sophisticated formulas need to be developed that align with human behavior.

 

Take that Siri.

 

 

 

10 comments

  1. I really enjoyed your post this week. I come from a very loud,sarcastic family and know for a fact how difficult it is to articulate that what I am saying to someone is not something offensive etc. that it is just sarcasm or a joke. I find it intriguing that social media forms are starting to adapt this as well. In a way I am surprise they haven’t prior but feel that censoring and filtering will aid this.

  2. adamsmea89 · ·

    I liked this post a lot! I had no idea an algorithm had been created to help detect this. I can imagine a very large portion of what is posed on social media platforms is sarcastic, so it would be very helpful for a computer to be able to detect this. Facebook monitoring chats for predatory behavior ties to some of the discussions we had in class about who’ job is it to police what is being said on social media. I think it is a good thing to have computers that can detect this type of conversation, because it would be impossible for a human to work through it all.

  3. I really enjoyed this post. The notion of sarcasm is an important one, the particularly as sometimes it is difficult to detect. I thought that the considerations for Facebook to explore an inappropriate conversation was interesting- especially the four criteria. I would think that this would mean that there may be cases that should be flagged and aren’t simply because of this criteria. Online sarcasm is also an interesting conversation for companies and their marketing. There has to be a way for companies to track sarcastic comments about their brands on social media site- especially to control negative feedback about their brand. It will be interesting to see how data analytics will have to change in order to pick up such commentary. While a computer may be good at ,some things, this notion of human interaction is something that cant be replaced by a machine.

  4. cattybradley · ·

    I think this is so important! I loved Del Harvey’s TED talk about monitoring on Twitter. Social media platforms have huge amounts of content to manage and if funny sarcastic jokes could be bypassed that would mean more time and attention for posts/content that raises real red flags.

  5. Great post! I think machines being able to detect human emotion has been a huge barrier in artificial intelligence. I was surprised to see how accurate the machine was in determining sarcasm (I feel like that’s a higher rate than most people understand sarcasm). It’ll be interesting to see if they can use a similar algorithm and apply it to different technologies. Like automated cars as they try to understand human behavior and how they would react in certain situations.

  6. finkbecca · ·

    Interesting topic! This relates perfectly to the TED talk we watched and some of the things we’ve been discussing in class. I feel like I usually have difficulty understanding sarcasm in text messages or on social media. It’s hard to tell the tone when something is written, so I feel more inclined to take it literally too. I am shocked that a mathematical algorithm was able to pick up on sarcasm 90 percent of the time, that’s impressive! Great post!

  7. fernaneq4 · ·

    Buzzfeed had a list (naturally) on common sarcastic phrases translated — https://www.buzzfeed.com/lukebailey/really-useful-list?utm_term=.ve4gZNOvZr#.amzPBvaNBE. It was funny because I forget how often that I’m actually sarcastic. There’s also research that shows sarcasm shows intelligence! (http://news.bitofnews.com/sarcastic-people-are-smarter/) I’m curious to see how the new algorithm will help companies like Facebook filter out what’s good and bad. In my own opinion, I don’t think a machine will ever be able to perfectly read through everything and label it morally correct or incorrect and we’ve seen examples of that in this class (googling “bad business hairstyles for women”). I wonder how they’ll be able to filter out things like dead baby jokes — “What is the difference between a baby and a onion? No one cries when you chop up the baby.” Obviously in the literal sense, this is terrible, but no one is being literal when it comes to dead baby jokes. Overall great post!

  8. katieInc_ · ·

    Great approach! When I read your title, I anticipated a post about how social media users perceive sarcasm. Instead, you took this in a much more interesting direction in terms of how technology can recognize sarcasm. This is clearly a huge opportunity to strengthen the Facebook’s effectiveness in monitoring inappropriate behavior as well as increase security at high volume events like concerts and sporting events. Out of all the different uses you mentioned, I am fascinated by how recognizing sarcasm can produce more accurate data for marketers to conduct research and analytics to better understand their target audiences. I am interested to see how this capability will be used in future marketing opportunities.

  9. skuchma215 · ·

    Great blog and grade-A gifs. Very interesting that Homeland Security has monitoring software to detect sarcasm, that’s something I would of never thought of. The fact that Yahoo! was able to detect sarcasm up to 90% is incredibly impressive. I think a lot of actual people fail to detect sarcasm through a text or email 100% of the time, I know I personally do. Besides governmental security monitoring and marketing analytics, I wonder what other applications sarcasm detection software could be used for. Maybe it could be implemented into AI software?

  10. Interesting post. I read a tongue-in-cheek movement a little while ago making the case for a form of punctuation that indicated sarcasm online. Of course, I just googled it and apparently there was more on the topic than I expected. https://en.wikipedia.org/wiki/Irony_punctuation

%d bloggers like this: