I've been thinking about blocking IPs using IPSec in Windows 2003 Server, programmatically.

Does anyone have any ideas? Recently, I've been receiving a lot of trackback spam and looking at it's occurance, it seems to be pretty ferocious. I decided to use IPSec to block certain IPs, so that they would not show up in the log files. However, it appears that the trackback spams originates from many IPs, and adding these IPs manually would be a tremendous task. Unfortunately, I've no idea where Microsoft places  it's IPSec API, so I can't program it at the moment.

Looking at the behaviour of the trackback spam, I'll say that it attempts to post lots of trackbacks within a minute. So, one way could be creating a trackback queue manager, and submit all the trackbacks there. The manager would analyze the trackbacks and determine if the contents of one trackback differs significantly from the previous trackback. If it does, then it's considered spam. I could be analyzing it by content, or by a neural network. If I decide to go the neural network way, I would be touching on a topic that I haven't touch on for at least 5 years.