Quote:
Originally Posted by Dephormation
Its all about context and volume of data.
A name alone - John Smith - might not identify an individual. With enough context it would certainly identify an individual.
You cannot allow a communication service provider (electronic or otherwise) to inspect all of your private communications and make a judgement for themselves about what is and isn't personal and private information before you've even had the chance to read it yourself.
You just can't do it.
|
Just did a quick word frequency test with TextSTAT, a simple word analysis program (free) on a Cable-Forum page from this afternoon.
Discarded any term that occurs less than 5 times ((I seem to remember phorm stating they do this)
Discarded the chaff (the, at, is etc)
Now it starts getting harder to narrow down.
Discarded generic words that would probably not be relevent (service, data, posted etc)
And i'm left with:
Phorm
Media
Dephormation
NTLVictim
Webwise
BT
roadrunner69
Paul
Delaney
Cable
software
VM
ISP
ICO
government
warescouse
Sky
video
davethejag
Joe
Phone
Google
Even from this, very limited test, and disregarding the url its possible to identify several people (at least by user name (Paul Delaney?), who's political views (at least with regard to privacy matters) are known.
Once the inevitable function creep comes who knows where it will lead.
I'll try the same thing on a more personal site when I get time.