Trolls clearly enjoy the magnification effect they can have by spreading hate speech to everyone subscribed to a group, but groups are also used by spammers, can also just have off-topic posts, or reply posts that probably aren't of general interest. It's hard to tell automatically whether a reply or a new post is likely to contribute to the conversation.
Pre-moderation of posts by volunteer moderators could make groups more useful and less prone to abuse. I've seen this model work well in other social media services, particularly for large groups.
If a group is configured to identify a particular set of moderators, then it could only boost posts if one of the moderators has liked the post, for example.
Trolls clearly enjoy the magnification effect they can have by spreading hate speech to everyone subscribed to a group, but groups are also used by spammers, can also just have off-topic posts, or reply posts that probably aren't of general interest. It's hard to tell automatically whether a reply or a new post is likely to contribute to the conversation.
Pre-moderation of posts by volunteer moderators could make groups more useful and less prone to abuse. I've seen this model work well in other social media services, particularly for large groups.
If a group is configured to identify a particular set of moderators, then it could only boost posts if one of the moderators has liked the post, for example.