Science Fiction Net, December 13th (Liu Yazhu) It is reported that with the rapid development of the Internet, a series of Internet illegal incidents have become a phenomenon that cannot be ignored in today’s society. Such as derivative human flesh search, traffic fraud and online violence and other behaviors that undermine public rules and violate the moral bottom line.
Recently, scientists have developed a new algorithm using two traditional criminological theories, daily activities and social infection, to build an artificial intelligence that can predict illegal events.
According to reports, this new type of AI can use chat data, including chat frequency, chat time, message recipients and other data to analyze and predict. The analysis results show that the accuracy rate of predicting illegal accounts is 84.85%, and the prediction of victim accounts can reach 85%. .
Moreover, this kind of AI can make use of powerful computer functions to perform “logical” calculations, to refine and find rules for scattered and distributed daily activity records. In the end, not only can it predict the time of illegal incidents in the online community in the next week, but it can also use AI to “catch people along the network line” to protect the network environment.