RE: virus: Nature of Information

Robin Faichney (r.j.faichney@stir.ac.uk)
Thu, 16 Oct 1997 13:40:25 +0100


> From: Sodom[SMTP:sodom@ma.ultranet.com]
>
[.........]

Remember to cut down quotes, please, folks.

> Right on Robin,
>
Thank you!

> Your last paragraph sums up the conclusion I recently came to.
> Patterns are both internal and external. I also came to the conclusion
> that since the brain lacks the ability to be "exact" that the pattern
> in
> our heads, that "matches" a pattern out there are basically just
> similar
> enough for us to use it.
>
Where a pattern is matched, that means we have two
instances of the same pattern, so it's always exact.
But remember patterns are abstractions. Say
pattern matching software is presented with two
scenes and set to match any corresponding patterns
in them. The parts that actually match (if any) will
generally be a very small proportion of the total
information in the scenes -- certainly if these are
photographs of actual concrete objects. A circle
matches another circle -- both are circles, and to
that extent it's an exact match. But obviously
many, many other aspects have to be ignored to
make that match.

> The more complex the pattern, the more likely
> it is to be confused with another pattern. Whatcha think?
>
You have to think about exactly what is a pattern.
Two things will be similar in some ways, different
in others. If the bits that are similar happen to be
the bits that are important to you, then it's an
exact match, otherwise it's not. But it does depend
on the context -- why are you comparing them,
what are you looking for?

Robin