Sarcasm Meter Calibration

[LongShortLong]

Your tax dollars at work

“Sarcasm has been a major hurdle to increasing the accuracy of sentiment analysis, especially on social media, since sarcasm relies heavily on vocal tones, facial expressions, and gestures that cannot be represented in text,” said Brian Kettler, a program manager in DARPA’s Information Innovation Office (I2O). “Recognizing sarcasm in textual online communication is no easy task as none of these cues are readily available.”

Researchers from the University of Central Florida working on DARPA’s Computational Simulation of Online Social Behavior (SocialSim) program are developing a solution to this challenge in the form of an AI-enabled “sarcasm detector.” The researchers have demonstrated an interpretable deep learning model that identifies words from input data – such as Tweets or online messages – that exhibit crucial cues for sarcasm, including sarcastic connotations or negative emotions. Using recurrent neural networks and attention mechanisms, the model tracks dependencies between the cue-words and then generates a classification score, indicating whether or not sarcasm is present.

https://www.darpa.mil/news-events/2021-05-06

TRG will never be the same

[pfluffenmeister]

Can't tell if serious.

[jono]

Can't tell if serious.

Sarcasm meter needs adjustment

[mfcf13]

Sarcasm meter needs adjustment

Context is important, too. If Bennymac posts "It's all just a big Plandemic!" you know he is being sarcastic. Whereas if Deebased posts the same thing, you know he is dead serious and not being sarcastic.


Similarly, if some Hillary doll fucker posts: "I've always thought the Achilles heal in Chinese society that will retard serious growth and innovation is their lack of real knowledge of how the rest of the world operates because the elites heavily control social media inside and outside the country. Very industrious people, smart as fuck, but they're flying blind when it comes to stuff like this, when experienced westerners could have virtually advised them."

...and then later posts "What have the Chinese ever invented?" you know the statement is not meant to be sarcastic and is just some eurocentric blather.

[jono]

^^ Sarcasm meter needs adjustment

[Tri-Ungulate]

Doug Pirhana

[old goat]

Your tax dollars at work

“Sarcasm has been a major hurdle to increasing the accuracy of sentiment analysis, especially on social media, since sarcasm relies heavily on vocal tones, facial expressions, and gestures that cannot be represented in text,” said Brian Kettler, a program manager in DARPA’s Information Innovation Office (I2O). “Recognizing sarcasm in textual online communication is no easy task as none of these cues are readily available.”

Researchers from the University of Central Florida working on DARPA’s Computational Simulation of Online Social Behavior (SocialSim) program are developing a solution to this challenge in the form of an AI-enabled “sarcasm detector.” The researchers have demonstrated an interpretable deep learning model that identifies words from input data – such as Tweets or online messages – that exhibit crucial cues for sarcasm, including sarcastic connotations or negative emotions. Using recurrent neural networks and attention mechanisms, the model tracks dependencies between the cue-words and then generates a classification score, indicating whether or not sarcasm is present.

https://www.darpa.mil/news-events/2021-05-06

TRG will never be the same

Doesn't work with Black people.

[jono]

Doug Pirhana

dramatic irony, metaphor, bathos, puns, parody, litotes and satire

I like a good hyperbolic understatement, myself, but the mastery of bathos isn't snowboarding. Hopefully the 400 years gets reduced.

[mattig]

God bless flying circus and those who can quote it verbatim years later. Really lets you know who your friends are or should be.

[jono]

Bump for RatG13 fans.

ATLANTA—Admitting that she was caught off guard by the resurgent threat, CDC director Rochelle Wolensky expressed alarm Wednesday after Googling “Covid cases” for the first time in weeks. “Holy shit, I haven’t been checking in on the coronavirus because I thought we’d turned a corner, but damn, this looks really bad,” said Wolensky, holding up her phone for the press to show the rapidly escalating graphs of caseloads and deaths displayed on the search engine before asking if anyone else had seen this. “And the more I’m looking at this, does the Johnson & Johnson one have a different effect on the delta variant than the mRNA ones? Jesus, I see why I’ve been avoiding it now. This is just super depressing.” Wolensky added that she would probably ignore hospitalization data for the next month and hope things leveled off before her upcoming vacation to Hawaii.