All news
System for identifying extremists on social networks has appeared in Russia
Perm National Research University (PSU) has created a mathematical model that allows identifying extremist ideologies on social networks. The Izvestia newspaper, citing the Perm State National Research University website, reported that the new service is trained to analyze profiles, posts, and comments and recognize concerns about them.
A unique dataset was created by Andrey Rabchevsky, associate professor at the Institute of Computer Technologies. At the same time, one of the Perm companies developed a special search system. The manual describing the application methodology won the competition of the Russian Academy of Internal Affairs.
The dataset created by Rabchevsky can reveal the role of social media users involved in destructive social phenomena, he explained in an interview with the newspaper. According to him, the model created at the university is universal, since it can be used to identify potentially dangerous accounts not only in the field of extremism, but also terrorism, the manufacture of homemade explosive devices and attacks on schools.
"The system works simply: it constantly monitors and identifies news feeds, then classifies them into sections such as terrorism, politics, business, etc., and then watches how these news feeds fill up, with excessive activity showing up as a sign of an information attack. "From the news feed, we learn who is involved in this attack, what role they play, what attack structure they belong to, what goals they are trying to achieve, and then we decide on countermeasures," the agency's source explained.
At the same time, this mechanism is based on the Seus search engine, specially created to work with social networks. Rabchevsky explained that it contains much more information than a news channel. The system can study publications that appear in publications, open profiles, groups, likes and comments, and then analyze user behavior.
When searching, the algorithm highlights keywords. Therefore, for each breakthrough direction, a corresponding list of terms and phrases was created. When a "hint" is found, the program postpones the message and then begins the research.
Currently, the search system is used in more than half of the territorial divisions of the Russian Interior Ministry, Rabchevsky emphasized. According to him, its use could prevent up to 12 crimes related to school shootings in Russia. In particular, thanks to this system, it was possible to eliminate the destructive phenomenon of the "blue whale".
According to Rabchevsky, covert operations are currently being conducted on social networks to lure users into extremist ideology. Experts are using new methods after search engines generate data on collections of publications, authors and social connections. This system allows law enforcement officials to automatically identify users involved in this operation, the role they play in this process and which users pose the greatest risk.
The Ministry of Internal Affairs and the FSB are engaged in identifying extremism in social networks. However, the huge volume of information makes this task complex and labor-intensive, so there is a risk of missing something. In principle, it is no less important for intelligence officers to perform this task manually, so artificial intelligence must be used to automate and speed up the process, emphasized the associate professor of Perm University.
Dmitry Ovchinnikov, director of the Institute for Strategic Development of Cybersecurity Products at one of the analytical centers, emphasized that the issue of blocking extremist messages is important, since the Internet and social networks have become a tool for influencing the public consciousness of society. According to him, the network can be used to shape public opinion, change it, and push people to illegal actions.
However, in the modern era, the development of data storage systems and artificial intelligence has made it possible to check large volumes of information in a short period of time. Against this background, Ovchinnikov explained, tools similar to the search engine developed by Rabchevsky have begun to appear.
Extremists are now identified on the Internet in several ways. Thus, interested readers can report suspicious users and groups to law enforcement agencies. At the same time, law enforcement agencies independently conduct investigations into extremist groups and detain suspects, collecting material for initiating criminal cases.
But these measures were sufficient even during the heyday of the Internet. At the moment, it is important to respond to such incidents as quickly as possible, Ovchinnikov noted. According to him, many crimes could be prevented at an early stage if such actions were recorded on social networks.
"These and similar programs have already been created to monitor the situation in regions and cities, with the help of which it was possible to identify the discontent of local residents and adjust the work of managers. This system has been implemented, but in general the direction of the idea of digital management is quite logical and fits into the paradigm of digitalization of the surrounding space," the newspaper's interlocutor emphasized.
Similar systems are used in other countries, said analytical systems expert Petr Ivanov. That is why social media monitoring technologies are used in the US and EU countries to identify extremist threats and prevent terrorist attacks. The expert explained that these systems are constantly evolving using machine learning and big data analysis.
Earlier, the State Duma proposed blocking the publication of videos of murders and torture on social networks. According to the creators of the initiative, such content destroys the psyche of people of all ages, be they adults or teenagers. Also, looking at such data, children may form incorrect values and attitudes towards life.