It looks like you're using an Ad Blocker.

Please white-list or disable AboveTopSecret.com in your ad-blocking tool.

Thank you.

 

Some features of ATS will be disabled while you continue to use an ad-blocker.

 

Researchers are Training Computers to Recognize Sarcasm on Social Media

page: 1
4

log in

join
share:

posted on Jan, 23 2016 @ 01:13 PM
link   

Twitter contains multitudes. On any given day you'll find earnest and passionate rumination, breaking news and analysis, silly hashtag games, horribly abusive idiots spewing hate and much, much more. Another constant across the platform is sarcastic reactions to all manner of events big and small. Indeed, when you're fully ensconced in the echo chamber that is Twitter, it can sometimes be hard to tell what's real and what's not. Fortunately, some researchers at Carnegie Mellon University have our backs: they're training computers to recognize sarcasm on Twitter, and they've had some solid success so far.

Source

PDF: Download

Extra: Rumor-detection software IDs disputed claims on Twitter

More linguistic detection software, great! (not) What exactly is the intent behind this? To help label or to help eliminate? Authors David Bamman and Noah A. Smith from CMU's School of Computer Science say its all about the context.

They wrote, “the relationship between author and audience is central for understanding the sarcasm phenomenon.” While this idea seems relative enough, the notion of one's audience begins to break down on social media websites like Twitter. Who exactly is your audience? Without any previous relationship, how does one express their thoughts intently?

You can attempt to imagine the audience you are trying to reach, if any, and hope for the best. But, as we see repeatedly, social media can feel like a train wreck in slow motion. Either you're the driver or somehow caught in the middle of a situation you never wanted to be in.


To properly test for sarcasm, the researchers built out a number of factors to test on. Individual tweets subjected to a number of factors, but the study also took into account details from the author's profile, historical content and details from that author's audience. It's a complicated bit of modeling, but testing on the tweet, its author, its audience and its response helped the researcher's sarcasm detector reach an 85 percent accuracy level. That's significantly higher than the 75 percent accuracy rate it hit when analyzing just the content of a tweet without additional factors included.

Again, why train computers to detect sarcasm? Its all about the future of automation and detection. Robots will need to better detect and understand the often undefinable intent or meaning behind someone's words. During security sweeps, you might want to hold back your sarcasm until their firmware gets upgraded.

More nefariously, the Secret Service had previously tried to detect sarcasm and false positives on Twitter in order to distinguish the difference between someone wanting to blow up an airplane with intent or for example, “I'm going to blow up this plane if it doesn’t get here on time!”

In my opinion, you are in idiot either way, but nobody deserves to be put on a list or be hassled by the man when airing their frustrations. Maybe five seconds of thought will help you develop a choice for better words and I would hope this same logic would prevail on ATS.

Think before you speak, think twice before you tweet. Thanks

(Tree House in a snow storm is bliss)


edit on 23-1-2016 by eisegesis because: (no reason given)




posted on Jan, 23 2016 @ 01:15 PM
link   
The sad thing is that they will likely be better at detecting sarcasm than some people are.

How bad is that?



posted on Jan, 23 2016 @ 01:19 PM
link   

originally posted by: ketsuko
The sad thing is that they will likely be better at detecting sarcasm than some people are.

How bad is that?

I agree, but I hope not.


edit on 23-1-2016 by eisegesis because: (no reason given)



posted on Jan, 23 2016 @ 03:17 PM
link   

Researchers are Training Computers to Recognize Sarcasm on Social Media


Wow. I'm sooooo impressed.



posted on Jan, 23 2016 @ 03:24 PM
link   

originally posted by: eisegesis

Twitter contains multitudes. On any given day you'll find earnest and passionate rumination, breaking news and analysis, silly hashtag games, horribly abusive idiots spewing hate and much, much more. Another constant across the platform is sarcastic reactions to all manner of events big and small. Indeed, when you're fully ensconced in the echo chamber that is Twitter, it can sometimes be hard to tell what's real and what's not. Fortunately, some researchers at Carnegie Mellon University have our backs: they're training computers to recognize sarcasm on Twitter, and they've had some solid success so far.

Source

PDF: Download

Extra: Rumor-detection software IDs disputed claims on Twitter

More linguistic detection software, great! (not) What exactly is the intent behind this? To help label or to help eliminate? Authors David Bamman and Noah A. Smith from CMU's School of Computer Science say its all about the context.

They wrote, “the relationship between author and audience is central for understanding the sarcasm phenomenon.” While this idea seems relative enough, the notion of one's audience begins to break down on social media websites like Twitter. Who exactly is your audience? Without any previous relationship, how does one express their thoughts intently?

You can attempt to imagine the audience you are trying to reach, if any, and hope for the best. But, as we see repeatedly, social media can feel like a train wreck in slow motion. Either you're the driver or somehow caught in the middle of a situation you never wanted to be in.


To properly test for sarcasm, the researchers built out a number of factors to test on. Individual tweets subjected to a number of factors, but the study also took into account details from the author's profile, historical content and details from that author's audience. It's a complicated bit of modeling, but testing on the tweet, its author, its audience and its response helped the researcher's sarcasm detector reach an 85 percent accuracy level. That's significantly higher than the 75 percent accuracy rate it hit when analyzing just the content of a tweet without additional factors included.

Again, why train computers to detect sarcasm? Its all about the future of automation and detection. Robots will need to better detect and understand the often undefinable intent or meaning behind someone's words. During security sweeps, you might want to hold back your sarcasm until their firmware gets upgraded.

More nefariously, the Secret Service had previously tried to detect sarcasm and false positives on Twitter in order to distinguish the difference between someone wanting to blow up an airplane with intent or for example, “I'm going to blow up this plane if it doesn’t get here on time!”

In my opinion, you are in idiot either way, but nobody deserves to be put on a list or be hassled by the man when airing their frustrations. Maybe five seconds of thought will help you develop a choice for better words and I would hope this same logic would prevail on ATS.

Think before you speak, think twice before you tweet. Thanks

(Tree House in a snow storm is bliss)



It's all about AI and the nuances that it will have to detect in order to become fully aware. Go watch ex machina....very interesting take on AI in that movie, and seems to be ahead of its time....almost visionary.



posted on Jan, 23 2016 @ 04:01 PM
link   
Great! So not only will the Skynet robots try to eradicate the human race they'll will be @ss holes while they do it.

"Hay M9 286, this human is trying to shoot me. I am leaking oil on the ground from fright. Please little human, do not shoot me with your tiny gun". -crushes human-



posted on Jan, 24 2016 @ 12:22 PM
link   
Nobody here is ever sarcastic, especially not me. None of us have anything to worry about.



posted on Jan, 24 2016 @ 05:39 PM
link   
a reply to: eisegesis

if they are training that means they are using AI, otherwise they are just programming



posted on Jan, 24 2016 @ 05:53 PM
link   
Looking at the statement from different contexts and historical usages and the consistency of the poster.

Mostly comes down to a percentage, which is good enough for thousands to millions of judgments.



posted on Jan, 24 2016 @ 06:22 PM
link   
Sarcasm depends on the self evident nature of the truth. Scary to think of an AI finding anything self evident. That could be a definition of sentience.

Sarcasm risks being petty and imprecise and can easily become a bad habit, IMHO.




top topics



 
4

log in

join