NATO secretary general comments on Russian military drillsWorld September 21, 21:34
NATO secretary general hails idea of deploying UN force in UkraineWorld September 21, 21:29
Russia ready to discuss alternative resolutions on UN mission to DonbassRussian Politics & Diplomacy September 21, 20:18
UN approves probe into Islamic State crimes in IraqWorld September 21, 20:10
Russia’s Alrosa mined all-time largest pink diamond in its historyBusiness & Economy September 21, 20:07
Russia submits Zvyagintsev’s film Loveless for OscarsSociety & Culture September 21, 19:16
Diplomat confirms Russia ready to support Iraq in fight against ISRussian Politics & Diplomacy September 21, 19:10
Russian, Syrian diplomats discuss cooperation within OPCWRussian Politics & Diplomacy September 21, 19:01
Putin talks to Russian Alisa voice assistant, inspects unmanned vehicle created by YandexScience & Space September 21, 18:33
MOSCOW, March 28. /TASS/. In teamwork with Belgian researches and Dutch police, Russian scientists have designed a computer program that can automatically spot child molesters in online chats, the National Research University at the Higher School of Economics reported on Monday.
The program designed by National Research University scientists is already used by Dutch police, its press release said. Analysis for research was based on a big number of texts. Texts were supplied by criminal police, while Russian specialists were responsible for the theoretical background of the system.
"We were to find out characteristics indicating that it was most probably a chat with a pedophile," the press service quoted a senior associate from the Higher School of Economics, Alexey Neznanov, as saying. All suspicions words and phrases were selected. Apart from the words themselves, the base included all possible options, such as abbreviations, mistakes and corruptions.
As a result of research, software was designed to make the process of search for pedophiles automatic.
Software was tested empirically; the results of research were described in a series of scientific articles.