The English language has numerous quirks and nuances that can trip up even fellow human beings, much less computers.  For instance, non-native speakers can often struggle to detect things such as sarcasm, but new research from the University of Central Florida suggests that computers may be getting more effective at doing so.

The researchers focused their attention on social media, due in large part to the strong use of the channel for customer service.  Sentiment analysis is increasingly used by companies to automate the identification of emotion in text.  While such analyses are reasonable at detecting broadly positive, negative, or neutral conversations, the new research has gone further into the nuances of language.

In common with other uses of machine learning, the researchers trained their system to spot patterns in text that identify sarcasm, while also teaching it to identify particular cue words within sequences that raise the chances of sarcasm being present.

“The presence of sarcasm in text is the main hindrance in the performance of sentiment analysis,” the researchers say. “Sarcasm isn’t always easy to identify in conversation, so you can imagine it’s pretty challenging for a computer program to do it and do it well. We developed an interpretable deep learning model using multi-head self-attention and gated recurrent units. The multi-head self-attention module aids in identifying crucial sarcastic cue-words from the input, and the recurrent units learn long-range dependencies between these cue-words to better classify the input text.”

Detecting sarcasm

Sarcasm has previously been regarded as a major obstacle to overcome in successful sentiment analysis, with the problem especially prominent on social media as sarcasm is heavily reliant on things such as facial expression, vocal tones, and gestures, none of which can easily be represented in textual communication.

“In face-to-face conversation, sarcasm can be identified effortlessly using facial expressions, gestures, and tone of the speaker,” the researchers conclude. “Detecting sarcasm in textual communication is not a trivial task as none of these cues are readily available. Especially with the explosion of internet usage, sarcasm detection in online communications from social networking platforms is much more challenging.”