Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
The best idea since raw toast.

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.

user:
pass:
register,


                     

Junk Email Analysis

If I only I had many, many samples of junk mail...
  (+2, -1)
(+2, -1)
  [vote for,
against]

Simple but may be effective. Correlate many different samples of junk mail (if only I had many samples of junk mail...) and find strong correlations. These could be common words, common word combinations, etc. The same sort of analysis used for speach recognition. If you wanted to get really fancy, use markov chains or something but even something simple might do the trick. I haven't seen anything like this floating around but if it was effective, I'd swear by it.
jeffie, Dec 30 2001

Here ya go … http://thespamletters.com/about.shtml
If I could save time in a bottle … I could throw away my gift lists. [reensure, Dec 30 2001]

A Plan for Spam http://www.paulgraham.com/spam.html
Bayesian spam filters are common nowadays (dunno if there were any in 2001) [quarl, Nov 09 2004]

Still stuff to be mined using this approach that Bayesian ignores http://www.gtoal.co...he_best_of_bot.html
You can find common substrings using an algorithm similar to LZ77 compression [gtoal, Mar 29 2006]

[link]






       ugh...that's _speech_ recognition.
jeffie, Dec 30 2001
  

       Are you suggesting the correlations could have some sort of meaning hitherto unseen? Messages from the gods? Government codes used to mystify and intrigue the general populace whilst dissemenating signals to suicide shoe bombers/undercover pedants/recipe card traders etc?   

       Do people actually read junk email? I shall look closely at the next one I recieve in the slim chance it may reveal the location of the microfilm!
benfrost, Dec 30 2001
  

       [jeffie], the Halfbakery lets you edit what you wrote. You can fix your typo and delete the annotation you added at the top.   

       «if only I had many samples of junk mail» Personally, I'd rather get sent less, not more.   

       [benfrost] I'm assuming that this idea is for a system which automatically removes junk email.
cp, Dec 30 2001
  

       For some reason, I don't get junk mail. I think earthlink blocks it somehow. (Not complaining, by any means.)
TeaTotal, Dec 30 2001
  

       Earthlink does...but I'd rather deal with it myself; no filter is perfect, and will delete something I DO want to see.   

       Besides, then I can be rude to them or forward them to mail fraud or the FBI...   

       And if you want a lot of samples of spam, I can set my filters up to forward them to you rather than delete them...
StarChaser, Dec 30 2001
  

       would someone please send me some junk - I have had no mail at all this weekend and feel so depressed.
po, Dec 30 2001
  

       Baked. E-mail filters already look for '!!!' or 'GET RICH QUICK', etc. If your e-mail program doesn't, make your own filter.
phoenix, Dec 30 2001
  
      
[annotate]
  


 

back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle