[Discuss] vnc
Bill Ricker
bill.n1vux at gmail.com
Thu Aug 28 19:07:21 EDT 2014
On Thu, Aug 28, 2014 at 5:33 PM, Edward Ned Harvey (blu)
<blu at nedharvey.com> wrote:
> Suppose there is a sentence, "your mother wears combat boots." Suppose an adversary knows this sentence. Then I think we'll all agree, the sentence has 0 bits of entropy, because entropy is a measure of how many guesses an adversary would need to make in order to correctly guess the unknown - and there is no unknown. Now suppose the adversary only knows "your __ther wears combat boots" and suppose the adversary knows the correct answer is either "fa" or "mo". Then I think we'll all agree the sentence has 1 bit of entropy,
add a +1 bit for m/\b ( combat | army) \W? boots \b /ix since Army
boots is the more popular cliché unless you know more Marines or
really liked the movie; and maybe +1.5 for space or hyphen or compound
word, and then there's capitals and jump-boots. And 'swims out to
troop ships'.
Suppose the adversary has a list of cliché sentences but doesn't know
which we've picked. How many bits entropy for one of them? (Add one
bit for cliché|not decision.)
Google Ngram shows 'your mother' is much more common than either
'mother wears' or 'mother (is) so', not unsurprisingly. Books are so
stuffy.
Another way to estimate bits is type into Google search and watch the
suggestions (in a logged-out, incognito/private session so no history
bubble).
y ..outube
your ... song lyrics
... mama jokes
m...ama jokes
m...omma jokes
your mo... mma jokes
your mot..her
...her was a hamster
...her should know
...her jokes
your mother ... was a hamster
... should know
... jokes
... wears army boots # is 4th.
your mother we...ars army boots
...ars combat boots movie
...ent to college
looks like 10-16 bits for either phrase, depending how much you credit
y,m,t,w and taking part or next suggestion.
--
Bill Ricker
bill.n1vux at gmail.com
https://www.linkedin.com/in/n1vux
More information about the Discuss
mailing list