Anti-spam issue

[AI-Avalon]

that the anti-spam program may be wrong. There was an incident in which another person (about 5 people) just made a normal statement, but everyone was judged to be spam and kicked. This is also slow, so I would appreciate it if you could solve it.

[clementwzk]

Maybe you spoke very quickly and the bot thought it was spam? @AI-Avalon

[AI-Avalon]

Hmm ... When 5 or more people were sending messages in 1 second increments, a warning was suddenly given to everyone who spoke on that channel, and even those who did not speak so much were detected as spamming their remarks more than the specified number of times. I was kicked.

[Skelmis]

Could possibly look into using some more advanced Antispam things? Or a classifier which is a bit smarter

[populated]

I'm super late.

It's because the filter works based on message age; it filters messages less than 15 seconds old regardless of who actually sent them. The check also does not compare the cached message's author to the author of the new incoming message; it only verifies if a message was sent within the last 15 seconds. Because of how this check works, if multiple people send messages rapidly (for instance, 5 different people each sending 2 messages within 15 seconds), someone could be warned even if they only sent that ONE message, and another person could be kicked even if they only sent one message.