What can we do proactively to prevent spammers from joining and posting on a Forem community?
For further actions, you may consider blocking this person and/or reporting abuse
What can we do proactively to prevent spammers from joining and posting on a Forem community?
For further actions, you may consider blocking this person and/or reporting abuse
Shital Kulkarni -
Olivia -
Benjamin Bilgehan -
Kryptobees -
Once suspended, michaeltharrington will not be able to comment or publish posts until their suspension is removed.
Once unsuspended, michaeltharrington will be able to comment and publish posts again.
Once unpublished, all posts by michaeltharrington will become hidden and only accessible to themselves.
If michaeltharrington is not suspended, they can still re-publish their posts from their dashboard.
Once unpublished, this post will become invisible to the public and only accessible to Michael Tharrington.
They can still re-publish the post if they are not suspended.
Thanks for keeping Forem Creators and Builders π± safe. Here is what you can do to flag michaeltharrington:
Unflagging michaeltharrington will restore default visibility to their posts.
Top comments (3)
I think itβs pretty hard to prevent spammers from joining as you would need to make the joining process harder so thatβs not worth the effort. This will make it harder for legit users as well.
Without being an expert on this, I guess generally spamming falls in the same domain as emails. So it should be possible to train a ML network to find and flag those posts and maybe even accounts. Those could be then reviewed by mods and flagged/unflagged which would train the network further. Maybe even a mail spam engine could be directly used although I suspect that they rely on the frequency of reports and sent mail to train the algorithm where we only have on post.
If this works, I think it is even better because spammers first do all the βworkβ to create an account and to post something just to find out that they got caught and blocked.
I think what we most need is some clarity over certain
if this then thatconditions where we can work to train a proper classifier.To this point, our different places have been pretty all-over-the-place and so the work we do is just too scattered for us to have high confidence in exactly what to tweak.
I feel like the most effective solution might be to do spam detection through a similar pattern to how we do human workβ marking something as spam and then letting someone confirm or invalidate the selection. That way at least there is clarity and parity in the process vs the black box.
Just wanna link to this post on DEV where I was gathering recommendations for text classification tools to fight spam.