The popularity of viral video has learned to predict using brain imaging. A new technique called neuroprediction, created by scientists at Stanford University, is described in PNAS.
Previous research by scientists has focused on the use of brain imaging to predict music sales, support crowdfunding, and general fundraising activities. As a result of the study, a structure called AIM (affect-integration-motivation) was created, which uses neural measures to predict user behavior and their willingness to spend money.
In a new work, researchers used the same system, but with changed settings – now its task was to predict how people allocate their time when watching a video.
The experiment was conducted on 36 volunteers whose brain was scanned using MRI while watching videos. The participants were shown 32 videos, and their main task was to stop playback for the first four to eight seconds.
At this time, the researchers monitored the brain activity of the participants in the experiment and compared the data with their reaction to a particular video. The scientists focused on four areas of the brain – the nucleus accumbens, anterior membrane, medial prefrontal cortex and posterior cingulate gyrus of the cortex.
The results showed that the determination of brain activity during the first four seconds of watching a video effectively predicts a person’s choice to continue watching the video and an emotional attitude towards it.
Activity in two specific areas of the brain was the most accurate metric in predicting which videos were most popular in terms of views on YouTube. Increased activity in the nucleus accumbens and decreased in the anterior envelope made it possible to predict the total viewing frequency better than any other metric studied, the authors of the study note.