Nov 14, 2019 10:55:00

Facebook reports that 3.2 billion false accounts were deleted in half a year

by

geralt

Facebook announced the 2019 Community Standards Enforcement Report on November 13, 2019 on service operations such as the removal of harmful content. Among them, it is reported that Facebook has deleted 3.2 billion false accounts between April and September 2019, more than double the previous year.

Community Standards Enforcement
https://transparency.facebook.com/community-standards-enforcement

Facebook removed 3.2 billion fake accounts between April and Sept.
https://www.cnbc.com/2019/11/13/facebook-removed-3point2-billion-fake-accounts-between-apr-and-sept.html

A fake account established by Facebook is an account registered with a different name or date of birth, which is created for the purpose of deceiving people clearly, those under the age of 13 or shared with others , Which avoids the registration requirements of the Terms of Service. Facebook reports that the number of fake accounts deleted from April to September 2019 was approximately 5% of Facebook's monthly active users.

The following graph summarizes the number of fake account deletions. The vertical axis represents the number of people (1B = 1 billion people), and the horizontal axis represents the quarter. Facebook reports that approximately 2.2 billion fake accounts were deleted from January to March 2019, a number comparable to the number of fake accounts deleted throughout 2018. Also, from April to September 2019, a total of 3.2 billion people, more than twice as many fake accounts compared to the same period in 2018 were deleted.

And the following graph is a bar graph showing the ratio of accounts that were discovered and deleted by Facebook before being reported (dark blue) and accounts that were deleted after being reported (light blue) among the fake accounts that were made. Facebook reports that most of them were discovered from July 2017 to September 2019 prior to the report.

In addition, Facebook has developed a detection technology that uses machine learning to detect and flag hate speech, saying that hate speech detection capabilities have increased. The graph below shows the number of hate speech deletions, which shows an increasing trend throughout the year. In July-September 2019, about 7 million hate speeches were deleted.

And if you look at the graph below that shows the percentage of deleted hate speech that Facebook found before being reported, you can see that the percentage that Facebook detects before reporting is also increasing year by year .

In addition, Facebook has set up a new category of harmful content, `` Suicide and Self-Injury '', and from April to September 2019, more than 1.6 million applicable content on Instagram. Revealed that it was deleted. According to Facebook, between April and September 2019, 4.5 million relevant content was deleted.

“We are pleased with advances in detection technology, but we recognize that these technologies are not perfect and that mistakes can occur,” said Guy Rosen, Facebook's Safety & Integrity division vice president. He said he will continue to invest in systems that detect hate speech and fake accounts.