Filter itself appeared in the app in September last year, but was arranged primitive.
The system simply deleted the comments with words and phrases included in the list offensive. The users could add their own words.
Now the system has become “smarter”. Using machine learning to detect unacceptable comments, given the context of the answer. The filter learns that, in future, to find more negative messages and reduce the number of false positives of the system. According to the website Wired, when artificial intelligence will mark the review, it blocks it for all users except the person who wrote it. So the attacker won’t even know that his message was concealed from the public eye.
The filter is now enabled by default — the previous had to be activated manually. If desired, you can disable it via settings Instagram. The ability to add your own words and phrases is still present. The system still works only in English, but over time the company plans to expand the list of languages.
Instagram also announced the spam filter — also on the basis of artificial intelligence. As it turned out, it appeared in service in October last year, but the company has officially told about it only now. This filter works in nine languages, including Russian.
© 2017, paradox. All rights reserved.
Russian President Vladimir Putin has set the stage for his state visit to China by…
С каждым днем онлайн-казино становятся все более популярными среди азартных игроков. И это вполне логично:…
Russian President Vladimir Putin has officially appointed Andrey Belousov as the new minister of defense.Belousov,…
Armenian police reportedly detained at least 38 protesters in the capital Yerevan on Tuesday, as…
The new Russian offensive in Ukraine’s eastern region of Kharkov puts President Vladimir Zelensky in…
Russian forces currently hold the upper hand in fighting near the city of Kharkov, Ukraine’s…