Usenet Binary Spam Filter
The most common method is to block the spamer's email, building blacklist for them. Or use some bot to send out cancel control messages to remove them. Click for more spam related resource. Shortages:
  • Spamer can fake their email addresses to get around the list easily.
  • Not proactive, always a step behind the spam.

    AUBA introduces a new spam filter method base on the signature of the binary contents. Named as UBSF, Usenet Binary Spam filter. Working for all binary usenet groups. Advantages,

  • Proactive, prevent it before it happend
  • Content based filter, spam can't get around it by faking header address.
  • Saving bandwidth
    How it works?
    Get the fingerprint(md5sum) of the binary conents, query AUBA's UBSF database. If it is a known spam, it will return string "SPAM", if not, it will return "OK". for example:
    http://archive.xusenet.com/ubsf.html?439cd6b58e60b328c3b1e4cc95be69cb (return OK)
http://archive.xusenet.com/ubsf.html?405fdda7fb56db2366758e8cc9d06ebf (known spam, return SPAM)
    Reject the post if it is a known SPAM. Filter programs for various NNTP servers will be available for download soon.
    What is UBSF database?
    UBSF database has been created and maintainted by AUBA, around 250k spams has been filted and block by AUBA currently. UBSF database is accepting contributed list from copyright owners who do not want their copyrighted material misposted by spammers or others.
    Simple, isn't it?

    Need some tech/faq writer's help, if you think you can contribute in this subject, please send your draft to "support@xusenet.com". Thanks. 

    Usage:
 http://archive.xusenet.com/ubsf.html?439cd6b58e60b328c3b1e4cc95be69ca

HTTP QUERY STRING must be an md5sum, it is 32 bytes hex.