They invent a sarcasm detector that Sheldon Cooper would need.

Who never wanted to have sarcasm detector? Or at least a small sign indicating that what you are saying is sarcasm. Speaking about this topic, to all the fans The Big Bang Theory One of his most iconic characters comes to mind: Sheldon Cooper. Perhaps he could use an algorithm that a group of scientists from the university has just presented. University of Groningen at the annual meeting of the Acoustical Society of America.

In fact, This is not the first sarcasm detector in history.. Since the advent of artificial intelligence, many algorithms have been developed that can detect when someone is saying something sarcastic. But, unfortunately, they tend to make a lot of mistakes.

The authors of the study just presented believe that this is due to the fact that their algorithms were based on only one parameter. However, they developed multimodal artificial intelligence algorithm, which is capable of both detecting feelings in speech and emotions in the manner of speech. All this together gives an answer to the possible presence of sarcasm.

What is a sarcasm detector used for?

In addition to warning our friends that what we said is not serious, the sarcasm detector can very useful apps. For example, those that were developed in the past were primarily intended to help virtual assistants to better understand the intent of user requests. They will also be useful for tools Content Analysis. The person watching the video no longer has to decide, for example, what audience it should be aimed at. A machine can do it. But for this it is important that it can also detect the presence of sarcastic content.

How did these scientists achieve this?

To develop their sarcasm detector, these scientists began by taking several recorded speeches, from which they extracted acoustic parameters such as rate, tone, or energy of speech. Transcription algorithms were then used that could convert the recorded message into text.

Thus, expressions associated with feelings were highlighted in speech and, in addition, they were placed emoticons corresponding to emotions associated with voice parameters. It is logical that this was first done with speeches of known intentionality, so machine learning algorithm.

Once trained, the algorithm was able to perform all of these steps: transcribe speech, detect expressions in words, and analyze feelings conveyed in voices. He kit All this was expressed in the presence or absence of sarcasm.

Other uses for this new sarcasm detector

We have already seen how the sarcasm detector is commonly used. However, since it is capable of analyzing both feelings and emotions, it is believed that it may have additional uses.

For example, it can be used in telemedicine to better understand patient messages and the emotions reflected in their voices. And also in marketing to fully understand product messages that users post online.

Without a doubt, the sarcasm detector is more than just a machine that would make Sheldon Cooper’s life easier. This is a machine that perhaps we will all use in the future. Time will show.