Computer science researchers at the University of Central Florida have developed a satire detector. Social media has become a major form of communication for individuals and for companies that market and sell their products and services.
Central Florida University Develops a Sarcasm Detector: That’s Cool!
Properly understanding and responding to customer feedback on Twitter, Facebook and other social media platforms is critical to success but is incredibly laborious. This is where sentiment analysis comes. The term refers to the automatic process of identifying emotion – either positive, negative or neutral – associated with the text.
Also, Check out Why Amazon Canceled 1 Month Prime Subscription!
While artificial intelligence refers to logical data analysis and feedback, emotion analysis is similar to identifying emotional communication correctly. The UCF team has developed a technique that accurately detects satire in a social media text.
How Does University of Central Florida Develops the Sarcasm Detector?
According to Assistant Professor of Engineering Evan Garibay, the main obstacle to the performance of sentiment analysis is the presence of sarcasm in the text. He said that it is not so easy to identify satire in conversation and therefore it is quite challenging to do it even for a computer program.
He stated that the team developed an interpretive intensive learning model using multi-head self-attention and gated recurrent units. Here, the multi-head self-attention module accomplishes identifying key sarcastic cue-words from the input.
The purpose of gated recurrent units is to learn long-range dependencies between these cue-words to better classify the input text. One of the team members, Ramya Akula, a computer science doctoral student, is working on it under a DARPA grant that supports computational simulation of the organization’s online social behavior program.
Brian Kettler, a program manager at DARPA’s Office of Information Innovation (I2O), believes that satire is a major hurdle when it comes to increasing the accuracy of sentiment analysis. This becomes even more important when the platform under consideration is social media.
He cited the reason that satire relies too much on vocal notes, facial expressions, and gestures that cannot be presented in the text. He further stated that identifying sarcasm in text online communication is not an easy task as none of these signs are readily available.
Also, Check out Apple Airpods 3 & Apple Music HiFi Release Updates!
How the UCF’s Sarcasm Detector Works?
The team taught computer models to find patterns that often indicate sarcasm and jointly taught the program to correctly select cue words in sequences that are more likely to indicate sarcasm.
He taught the model to do this by feeding large data sets and then checking their accuracy. The team also included computer science doctoral student Ramya Akula. Akula said, “In face-to-face conversation, satire can be easily identified using facial expressions, gestures, and the tone of the speaker.”
He said, “Detecting satire in text communication is no small feat as none of these signs are readily available. Especially with the explosion of internet usage, detecting sarcasm in online communication from social networking platforms Far more challenging. “