📜 ⬆️ ⬇️

Twitter changes user policy

Twitter announced a policy change to tackle abuse. The company points out that it continues to improve its platform on which users can safely interact with the world as a whole.

There are only two fundamental changes. One is related to the prohibited content, and the other to the reaction to certain violations of the rules.

Previously, the ban was limited only to "direct, specific threats of violence against others", but now extends to "threats of violence against other persons or the promotion of violence against others." The company says that the previous policy was too narrow and limited the ability to act on certain types of threatening behavior. In some cases, the service required users to delete content or confirm a phone number. Now the support team gets the ability to block suspicious accounts for a certain time.

Twitter is also testing the automatic detection of abusive messages and limiting their scope. This feature will take into account a wide range of parameters, ranging from the correlation of messages and ending with the age of the account. This feature, according to the social network, will not affect the display of content of hidden accounts, and will also help to limit the potential harm from the offensive content of the posted messages.
')
Back in February, leaked to the press, a note from Twitter CEO Dick Costolo (Dick Costolo) talked about the long-term disregard for the issues raised.
“We suck at it for years and we’ve been sucking at it for years. It’s not a secret. Addressing simple trolling issues that they face every day. ”

On April 16, in the Washington Post column, chief lawyer of the company, Wilaya Gadde, wrote about attempts to stop the violence, while maintaining freedom of speech.

The ultimate goal of such changes is to provide secure access to information for the widest possible range of users. Twitter representatives promise that the changes will not remain without control and evaluation.
Twitter_locked


The Techcrunch website provides a study published by Cornell University, analyzing a similar approach. You can also read my topic on a similar topic ( Method of identifying "trolls" in online communities using the example of Q & AC ).

Source: https://habr.com/ru/post/289868/


All Articles