Actually, I've thought about this, but have expanded the idea. After all, you don't want the poster having complete control over how it gets tagged do you?
So, I've decided to build smart system that will learn from a members' posts and tag them appropriately and apply properties to the post text, much in the way downvoted text gets hidden on blogs. It's been on in the background learning for about three months now and it seems to have come close to being able to go live. Here's some of the latest flags it's applied to the users:
Taffy: Advertisement sprinkled with derision
Text Decoration: Flashing multicolored text with all words being linked to his site.
Example:
I'd say you're a hobknocker and you need to purchase my jetting kit.
Bushmechanic: Over your head
Text Decoration: Font size will be reduced to -20, making the text tiny and the post impossible to read to save space, since you're probably not going to be able to wrap your head around it anyway.
Example:
You need to take the left-handed smoke sifter and attach it to the flux capacitor, thereby creating a rip in the time-space continuum and effectively allowing your Berg to travel through time.
Schwim: Of no consequence
Text Decoration: Font color will be changed to match the background color of the post box, effectively making the post invisible.
Example:
There's more in the database, but I've not yet implemented "tone it the hell down" feature, that softens the blow for the target of the system. Currently you have to have really tough skin to handle the psychological impact of being auto-moderated like this.
thanks,
json