Online testing of an artificial intelligence voice recognition system that can monitor and identify pornographic content began on Sunday, according to Science and Technology Daily.
Helped by the voiceprint recognition method, the Alibaba voice recognition system can identify multiple languages such as Chinese, Japanese, English and Russian, as well as dialects from Chinese provinces such as Hunan, Hubei, Henan, Sichuan and Guangdong.
Transforming voice into scripts, the system compares the scripts with keywords in its lexicon and anti-spam audio models which were also developed by China's tech giant Alibaba.
The lexicon and anti-spam audio models collect tens of thousands of pornographic words with the same or similar pronunciations, Alibaba told Xinhua.
The system monitors both online and offline voice files.
The multiple languages and dialect recognition ability need to be trained like a robot. For example, the system's Cantonese recognition ability was cultivated by watching TV series.
The system is scheduled to be put into operation in September.