Peter Diamandis, founder and executive chairman of the XPRIZE Foundation, talks about the pros and cons of artificial intelligence on “Craman Countdown.”
A team of Dutch university researchers Artificial intelligence (AI) Platforms that can recognize sarcasm, according to a new report.
The Guardian newspaper reported The research was presented Thursday at a meeting of the Acoustical Society of America and the Acoustical Society of Canada in Ottawa, Canada, the researchers said Thursday. At the event, Ph.D. student Xiyuan Gao said the research team used video clips from American sitcoms such as “Friends” and “The Big Bang Theory” as well as other text and audio to train the neural network. The content was used.
This study utilized a database called Multimodal Sarcasm Detection Dataset (MUStARD). Another research team in the US and Singapore was labeling and annotating specific content as to whether or not it was present as part of a push to build their own irony detectors. .
After training the AI model on the data, the researchers said it was able to detect sarcasm in interactions that the researchers had not labeled nearly 75% of the time. They added that subsequent laboratory studies using synthetic data further improved that level of accuracy, but the results of those studies have not yet been published, the Guardian writes.
AI is like a “tsunami'' that will hit the workforce, says IMF chief
A team of researchers from the University of Groningen has announced the creation of an AI model that can detect sarcasm. (St. Petersburg)
Some of the content from TV program The database used to train the AI model on sarcasm includes scenes from “The Big Bang Theory'' and “Friends,'' in which Sheldon observes his friend and roommate Leonard's failed attempt to escape from a locked room. It was. Chandler, Joey, Ross, and Rachel are not busy assembling furniture.
Matt Koller, a researcher at the Speech Technology Laboratory at the University of Groningen, told the magazine: “We are able to recognize sarcasm in a reliable way, and we want to develop this further.” Told. “We want to see how far we can push it.”
Worried about losing your job to AI? Experts point out real threats
A scene from The Big Bang Theory involving Leonard Hofstadter (Johnny Galecki) and Sheldon Cooper (Jim Parsons) was used in the AI Sarcasm training dataset. (Michael Yarish/CBS via Getty Images)
Shekhar Nayak, another member of the research project, said the team's methods and findings help AI assistants interact more easily with human speakers by detecting negativity or hostility in a speaker's voice. He said there is a possibility that it could be done.
Gao noted that incorporating visual cues into AI tools' training data could help them detect more sarcasm conveyed through facial expressions, such as raised eyebrows or fake smiles.
The future of love: Bumble founder says AI could do the dating for you
Scenes from the sitcom Friends starring Matthew Perry as Chandler Bing, Matt LeBlanc as Joey Tribbiani, and David Schwimmer as Ross Geller were included in the AI Sarcasm Detector training dataset. Ta. (Danny Feld/NBCU Photobank/Getty Images)
The interest in AI-powered sarcasm detection demonstrated by the University of Groningen research team and the researchers who compiled the sarcasm-annotated MUStARD content database follows similar lines of research conducted by the U.S. government. is following. US Department of Defense in recent years.
Researchers at the Department of Defense's Defense Advanced Research Projects Agency (DARPA) and the University of Central Florida used DARPA's SocialSim program to create an AI model that can classify whether a piece of text is: Developed. social media posts Or the text message contains sarcasm.
CLICK HERE TO GET FOX BUSINESS ON THE GO
“The team demonstrated the effectiveness of their approach by achieving state-of-the-art results on multiple datasets from social networking platforms and online media,” DARPA wrote in 2021. “The model was able to predict sarcasm well, achieving near-perfect prediction” Sarcasm detection score on the measure Twitter benchmark dataset We also provide state-of-the-art results on four other important datasets. ”
“Accurately detecting sarcasm in text is only one part of the development of these simulation capabilities, as human communication uses highly complex and diverse linguistic techniques. “Knowing when it will be used is valuable in teaching models what human communication is like, and then simulating the future direction of online content.”
