Textual analysis

Tim May writes:
>Imagine what can be done with word and phrase frequency analysis, with
>examination of punctuation styles (e.g., some people use _this_ for
>emphasis while others use *this*), and so on. Entropy measures, etc.

   I know for sure that Soviet KGB did a lot of work in graphology and 
kept samples of print of every typewriter there was in the country.
<not that it helped them ;) >

   It might be easy to write a program that would randomly modify spacing,
indentations, punctuation styles, spelling, replace words with random
synonyms, reorder words in phrases, etc.  It can eliminate most of the
clues, excluding the concepts.
You will have to compromise between the accuracy of the message and its
privacy protection, but it is still something...

Alexander Chislenko

Tim May writes:

>Imagine what can be done with word and phrase frequency analysis, with
>examination of punctuation styles (e.g., some people use _this_ for
>emphasis while others use *this*), and so on. Entropy measures, etc.

I believe that such programs already exist. They are used to identify
unknown authors of famous documents.  I believe some of Shakespeare's
work was under close scrutiny a number of years ago, as were the
Fedaralist papers.

There was an article in Byte about three or four years ago all about
this kind of technology.

Thug