Content Filtering

I was thinking about an effective filtering mechanism for the content of interest: pre-moderation, karma, etc ...

And a simple idea came to mind: why not use Bayesian filtering for topics? In theory, it should work even better for this purpose than for spam filtering. In addition, if the loss of email due to a filter error can be very expensive, then skipping Habratopic is absolutely not critical.

I don’t know, maybe the idea is crazy ... I didn’t really think about it yet. :)