ANON: Textual analysis

> This reveals a minor and probably obvious weakness of pseudonyms--writing
> styles.

We probably need "rephrasing remailers" which do some rudimentary
grammar parsing on input text, and randomly substitute equivalent
constructs such as switching active/passive voice, synonyms, changing
the word order where it is insignificant, joining/splitting sentences,
etc.  Anyone here have any experience in NLP (natural language processing),
specifically parsing english?

A possible start would be to look at "grammar checker" programs that 
check for various grammatical mistakes/misusages and suggest improvements.

Another starting point is language translation software.  After your text
has been translated automatically to spanish -> french -> german -> english,
not much of the orignal style will remain.  Hopefully, enough meaning
will be preserved to allow understanding.

Are there any public domain programs that do one of the above?

One constraint on these is that the message must be present in clear text,
so that it must be the last remailer in the chain.

> examination of punctuation styles (e.g., some people use _this_ for
> emphasis while others use *this*)

This could be alleviated by using a standard markup format, such as
MIME RichText, or the simpler markup convention recently proposed
on the mime list.


--
Yanek Martinson
yanek@novavax.nova.edu

Yanek Martinson stated in the last message:
>From cypherpunks-request@toad.com  Tue Mar  2 23:06:23 1993
>Subject: ANON: Textual analysis
>Reply-To: yanek@novavax.nova.edu

>> This reveals a minor and probably obvious weakness of pseudonyms--writing
>> styles.
>
>We probably need "rephrasing remailers" which do some rudimentary
>grammar parsing on input text, and randomly substitute equivalent
>constructs such as switching active/passive voice, synonyms, changing
>the word order where it is insignificant, joining/splitting sentences,

That is insane... remailers are not the place where this kind of work
should be taking place. If anything the user should run his document
that (s)he wants stripped of grammer clues with software designed for
that purpose on on his/her own machine.  This way, the user can see
immediate results.  They don't have to wait anywhere from 1 minute to
a few days to see the results of the "rephrasing remailer".

-- 
+==== Internet: babani@cs.buffalo.edu ===+======== Amateur-Radio: N2LYC ======+
!      Bitnet: V078LNGT@ubvms.BITNET     |        UUCP: rutgers!ub!babani     !
! Alternate: an173@cleveland.freenet.edu | Plsure dpnds on the othrs prmison. !
+==  PGP key available. (If you don't know what a PGP key is... find out!)  ==+

>We probably need "rephrasing remailers" which do some rudimentary

	 That is insane... remailers are not the place where this kind of work
	 should be taking place. If anything the user should run his document

Hardly insane.  If the rephrasing software were available for
Connection Machines and nothing else, then one would very much want a
remailing server that would rewrite phrases for you.  It'd be
wonderful if it worked on Joe Schmoe's 286 box, but it's pretty
unlikely.

Until it does, using such capabilities built into remailers would be
an improvement over the current situation.  

Note that I'm not making any claims that the technology is possible
any time soon. 

dean

E. Dean Tribble stated in the last message:
>From memexis!tribble@uunet.uu.net Fri Mar  5 03:44:47 1993
>Date: Thu, 4 Mar 93 10:31:17 PST
>From: memexis!tribble@uunet.uu.net (E. Dean Tribble)
>To: uunet!cs.Buffalo.EDU!babani@uunet.uu.net
>Subject: ANON: Textual analysis
>
>	 >We probably need "rephrasing remailers" which do some rudimentary
>
>	 That is insane... remailers are not the place where this kind of work
>	 should be taking place. If anything the user should run his document
>
>Hardly insane.  If the rephrasing software were available for
>Connection Machines and nothing else, then one would very much want a
>remailing server that would rewrite phrases for you.  It'd be
>wonderful if it worked on Joe Schmoe's 286 box, but it's pretty
>unlikely.

I was under the impression that such software could be developed for
Joe Schmoe's 286 box or even Joe Schmoe's 386 box or (if Joe is rich)
his 486 box.  The power of these computers shouldn't be underestimated.

The first step is to develop the programs. Are there any programs
redily available for testing and/or analysis on any type of platform?

-- 
+==== Internet: babani@cs.buffalo.edu ===+======== Amateur-Radio: N2LYC ======+
!      Bitnet: V078LNGT@ubvms.BITNET     |        UUCP: rutgers!ub!babani     !
! Alternate: an173@cleveland.freenet.edu | Plsure dpnds on the othrs prmison. !
+==  PGP key available. (If you don't know what a PGP key is... find out!)  ==+