ietf-asrg
[Top] [All Lists]

Re: [Asrg] Requirements for gathering statistics

2003-03-24 09:08:06
Kee Hinckley <nazgul(_at_)somewhere(_dot_)com> wrote:
I assume that we understand that these numbers are going to be very 
inaccurate?  Because if they were accurate, then presumably we'd know 
how to identify spam and we wouldn't be having this conversation.

  I'm not quite sure what you're getting at.

- The numbers will be precise
  byte or message counts are easy to calculate correctly

- The numbers may be accurate
  they will count exactly the messages which the ISP's believe to be
  spam, and not any other messages

- The criteria for determining *which* messages are spam is vague,
  imprecise, ill-defined, and subject to personal bias


  However, by measuring data from from biased systems, we can get a
good indication of what those biases may be, how extensive they are,
and even discover the shape of the non-biased systems.

  Pollsters use these methods all the time.  They call people, collect
precise, accurate data about personal biases.  The people may lie, but
with enough data, that bias may be accounted for.

  I'm not saying the numbers will be perfect, or perfectly collected.
I'm coming at it from a method used in physics: A small step forward
is better than sitting still.

  What amazes me is the number of people who would rather do nothing
than use an imperfect system, and the number of people who fight
against something they agree with, because they think it doesn't go
far enough.

  Can we please learn how to walk before we start running?

  Alan DeKok.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg



<Prev in Thread] Current Thread [Next in Thread>