A tip for a scammer: they learned to identify potential victims using their avatars

Scientists have developed a method for assessing the security of personal accounts

What is a person’s psychological defense mechanism, his avatar on social networks will tell you. Researchers from the St. Petersburg Federal Research Center of the Russian Academy of Sciences, together with researchers from the Russian Presidential Academy of National Economy and Public Administration, trained a neural network to find this relationship.

Photo: freepik.com

Social networks — This is a useful tool for communication, dating, and doing business. But scammers often use them, based on a number of criteria, choosing from thousands of users the most gullible citizens who are poorly protected psychologically.

As the Russian Academy of Sciences reported, scientists from St. Petersburg decided to analyze citizens’ accounts for vulnerability to cyber attacks.

According to Fyodor Bushmelev, an employee of the Laboratory of Theoretical and Interdisciplinary Problems of Informatics of St. Petersburg Federal Research Center of the Russian Academy of Sciences, in order to gain the trust of a potential victim, cybercriminals carefully analyze what she writes and what photos she posts online. It is also important what kind of avatar the account owner displays on his social network.

If you get ahead of the scammer and check your page in advance for psychological vulnerability, you can, theoretically, avoid cyber attacks in the future.

< p>In order to carry out “diagnostics” for each of the tens of millions of users. using his data, scientists trained a neural network to do this. To do this, they used the same patterns (schemes-images) that attackers usually use.

For example, at the first stage, scientists conducted a study using the psychological test “Life Style Index” to identify the levels of expression of psychological defense mechanisms among respondents according to their emotional status (joy, sadness, anxiety). 300 people volunteered for this study. Then the neural networks were “showed” avatars of volunteers to  train them to establish correspondences between the owners’ already known lifestyles and the features of photographs.

 The program already recognizes the position of the user’s head, eyes, and facial expression.  In the future, scientists want to add more features to its database, using, for example, the position of the body in the frame, pose, objects and other features.

By the way, neural networks in the future may help HR specialists to “get to know” people in advance. with potential employees of the company, assessing their activity by avatar.