Properly understanding and responding to customer feedback on Twitter, Facebook, and other social media platforms is critical to success, but it is incredibly labor-intensive.
There comes an analysis of opinions. The term refers to an automated process that identifies a feeling associated with a text — either positive, negative, or neutral. Although artificial intelligence refers to logical data analysis and response, emotional analysis resembles the correct identification of emotional communication. The UCF team developed a technique that accurately detects sarcasm in social media text.
The group’s findings were recently published in Entropy.
Effectively, the team taught the computer model to look for patterns that often show sarcasm, and combined it to teach the program to correctly select clues in words in series that are more likely to refer to sarcasm. They taught the model to do this by feeding it large data sets and then checking its accuracy.
“The presence of sarcasm in the text is the biggest hurdle in performing emotional analysis,” says assistant professor of technology Ivan Garibay ’00MS’ 04PhD.
“Sarcasm is not always easy to identify in a conversation, so you can imagine a computer program quite challenging to do it and do it well. We developed an interpretable in-depth learning model that uses multi-faceted self-attention and a fenced repetitive unit. hints from the word input, and repetitive units learn long-range dependencies between these hints to better classify input text. “
The team, which includes Ramya Akula, a PhD student in computer science, began addressing this issue under a DARPA grant that supports the computational simulation of an organization’s online social behavior.
“Sarcasm has been a major barrier to increasing the accuracy of emotional analysis, especially on social media, as sarcasm is heavily dependent on sounds, expressions, and gestures that cannot be represented in the text,” says Brian Kettler, Director of DARPA’s Information Innovation Office (I2O). “Identifying sarcasm in text messaging is not easy because none of these tips are readily available.”
This is one of the challenges that Garibay’s Complex Adaptive Systems Lab (CASL) is exploring. CASL is a multidisciplinary research group dedicated to the study of complex phenomena such as the global economy, the global information environment, innovation ecosystems, sustainability, and social and cultural dynamics and evolution.
CASL researchers study these problems through, among other things, data science, network science, complexity science, cognitive science, machine learning, in-depth learning, social sciences, group knowledge.
“In face-to-face conversation, sarcasm can be easily identified using expressions, gestures, and speaker tone,” Akula says. “Detecting sarcasm in text messaging is not a trivial task because none of these tips are readily available. Especially in the context of the explosion of Internet use, detecting sarcasm in online communication from social networking platforms is much more challenging.”
Garibay is an assistant professor of production technology and management systems. He holds several degrees, including a Ph.D. from UCF. Garibay is the director of UCF’s CASL artificial intelligence and Big Data Initiative as well as a master’s program in data processing.
Her research interests include complex systems, agent-based models, social media information dynamics, artificial intelligence, and machine learning. He holds more than 75 peer-reviewed papers and more than $ 9.5 million in funding from several National Agencies.
Source: The Nordic Page