Scientists have developed an artificial intelligence system that predicts to what extent a relationship would keep going based of voice patterns, according to a research published in the scientific journal PLOS One.
The machine learning algorithm, created by scientists from the University of Southern California, utilizes the tone and intensity of your voice – highlights like pitch, variation in pitch and intonation– to foresee what will be the possible result of your relationship.
In spite of the fact that the system doesn’t consider the spoken words itself (literal meaning) or the non-verbal communication of the speaker, it still out beat human predictors in a study that was done utilizing vocal recordings of 134 couples in unsettled relations with a precision of about 80%.
To begin with, the scientists utilized therapy session recordings of these couples to get their vocal characteristics.
At that point, they put these characteristics into the machine learning system, alongside data on the duration of the relationships.
The machine learning algorithm was then programmed to learn the relationship between the vocal highlights, examining who talked when, to what extent and the sound of their voices.
Examining the examples encoded in the way that they talked, the machine learning algorithm figured out how to anticipate the result of the couples’ relationship with very high precision.
Indeed, it was far better than specialists who evaluated the vocal tone as well as the words utilized by the couples.
To be more particular, against the Artificial Intelligence’s accuracy rate of 79.3% , the specialists could accurately predict 75.6% of the cases.
While this system could give couples opportunity to understand that their tone could be one reason for their relationships not ending great, it likewise gives bits of knowledge into how our method for talking could contain more data about our hidden sentiments than we know about.