Tech

Twitch will use machine learning to improve the security of its chats

Over the past few months we have watched Twitch begin his personal crusade to establish stronger measures against so-called “hate raids”, a problem that is increasingly present on the platform, and that led streamers and Twitch users to carry out the first strike that the platform has experienced.

Thus, although the last measure adopted was to limit participation in them only to users who have their Twitch accounts validated via email and / or mobile phone, now Twitch continues to advance its efforts to reduce bullying with a new tool that uses machine learning, with which they hope to detect those who may be trying to evade these new rules.

The new tool, called Suspicious user detection (Suspicious User Detection), it will be able to autonomously identify users as “likely” or “possible” people to be involved in trying to evade the channel’s bans or bans of a streamer. The machine learning model that drives the tool identifies potential evaders by evaluating aspects such as their behavior and the characteristics of their account, comparing that information with the accounts that have been banned from the channels of their platform.

Twitch Suspicious User Detection

Operating as a silent ban, messages from these suspicious users will not be sent to the chat (although they will not have any notification about it) being only the streamers themselves and their mods able to see them. In this way, channel owners will be able to choose to use the new monitoring tool to follow the behavior of these users, offering them better knowledge before having to resort to expulsion.

As Twitch itself has shared, this new user detection function will be active by default from its implementation, but the channels will be able to modify or deactivate the tool if they wish. Additionally, although the machine does not detect some potential suspicious users, the monitoring tool can also be used manually.

While it remains to be seen how effective this tool can be, it is certainly a big step in trying to avoid these toxic practices within this community.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *