A round table discussion “Countering the ideology of terrorism in the media: the potential of artificial intelligence (AI) for detecting radical content” was held in Moscow.
The discussion was attended by representatives of international organisations, government agencies, the media, the expert community, and former law enforcement officials.
Roman Zezyulia, Deputy Director of the Department for State Support of Media Development at the Russian Ministry of Digital Development, Communications and Mass Media, thanked the organiser, the International Academy of Television and Radio and the Russian Ministry of Digital Development, Communications and Mass Media:
“The topics change, but the format of the round table remains the same – intimacy, the opportunity to speak and be heard – this is the strength of such events,” he said.
Vladimir Kuznetsov, Director of the UN Information Centre in Moscow, emphasised that the use of artificial intelligence in the fight against terrorism is a complex but important issue that can unite states.
“The nature of the discussion and the goals set by its participants are undoubtedly aimed at bringing people together. […] I remain optimistic and truly hope that common sense will prevail. But for it to prevail, everyone must join forces,” he said.
Yulia Nefedonova, Assistant Professor at the Faculty of Journalism at Moscow State University (MSU), presented a study showing how technology recognises the emotional tone of messages and explained why it is useful to apply artificial intelligence.
“Texts distributed through radical channels often have a very professional informational and psychological impact. Users are far from immediately able to recognise the problem of radical content and distinguish it from neutral content, and sometimes they cannot do so at all,” she said.
Yulia Nefedonova noted that the volume of information no longer allows us to rely solely on manual monitoring.
“We started using the BERT neural network to identify signs of social conflict in media content. The model was unable to classify the types of conflicts, but it coped successfully with the task of binary analysis, with an accuracy of about 70 per cent,” she said.
In turn, Maria Anikina, Associate Professor at the Faculty of Journalism at Moscow State University, continued the theme, emphasising the importance of automated technologies in information analysis. She spoke about her research, which was based on the study of more than 300,000 social media posts over a month.
According to her, modern tools and programming environments make it possible to expand the capabilities of analysing large amounts of text information compared to traditional manual research.
(TV BRICS)
