According to a new paper in the International Journal of Computational Science and Engineering, artificial intelligence and text mining techniques may be used to identify social media users classified as paranoid based on their set of posts on Twitter.
“This paper proposes an architecture taking advantage of artificial intelligence and text mining techniques in order to: 1) detect paranoid people by classifying their set of tweets into two classes (paranoid/not-paranoid); 2) ensure the surveillance of these people by classifying their tweets about COVID-19 into two classes (person with normal behaviour, person with inappropriate behaviour),” the study’s authors explained in their findings.
For the study, a team of researchers at Universities in Tunisia and Saudi Arabia set out to examine the behavior of users on Twitter during the COVID-19 pandemic to detect personality disorders related to paranoia.
The findings underscore the significance of people who are driven by a mistrust of authority and easily swayed by disinformation.
“Ultimately, the team’s analysis of Twitter users discussing COVID-19 could allow them to find people who may be suffering unduly and may be entering a personal crisis,” a news release of the findings read.
“In other words, the tools they discuss could be used as a proxy diagnostic that could allow qualified professionals to offer an appropriate intervention for patients with paranoia.”
