Spam Filter

About a year ago, I started to set my filter rules in webmail. The spam filter has successfully helped to eliminate over 3000 junk email messages, average 10 spam mesages are caught per day. Still not good enough as I have about 50 junk email a day. I need to make tighter control in filtering.

Increasingly, spammers send out HTML mails with images containing adveritizement to avoid spam messages being filtered by text-based rules. Nothing on Earth can filter this kind of spam mail.

Here is the hit rates of my text-based filtering rules :

No. of hits String in message header

17 Award
91 Free
36 Application
120 Cheap
138 Free
34 Fw:
27 Fwd:
8 Male
53 N/A
11 Notice
284 ci (target word is cialis)
144 med (taget word is medicine)
111 online (target word is online porn)
48 order
51 pharmacy
111 price
24 shop
12 size (target word is penis size)
60 software
2 strength
99 weight

No. of hits String in message body
97 Generic
172 cheap
107 cialis
1022 click
3 good price
87 loan
106 mort (taget word is mortgage)
85 rates
82 unsubscribe
80 viagra

No comments: