Read an article not too long ago in The Economist about this new concept for a database. Apparently, some fella named Hempel, back in the 1940s, took the statement "all ravens are black" and derived its negative equivalent by saying "all non-black objects are non-ravens." Some Yale (had to include that for you, Dad) computer scientists have taken this line of thought and are attempting to apply it to the problem of protecting sensitive data.
I'd try and paraphrase but those Economist writers (whomever they are--why don't they list authors' names in that magazine?) say it best:
"The idea is to create a negative database. Instead of containing the information of interest, such a database would contain everything except that information."
Try and wrap your head around that crazy-ass shit.
8 comments:
Aren't you the well-read, intellectual? That is a bizarre concept -- and no -- I can't get my mind around it!
Also -- I like the photo!
thanks for the comment, daniel fan... still don't know WHO the hell you are.. it's nice to know i've got a fan out there tho! by the way, you don't need a comma after "well-read."
DAMN IT -- that comma is a typo by the way :(
WHAT?! That is so bizarre. It seems like something out of a Thomas Pynchon novel or something and like my worst nightmare.
I think Daniel has an admirer...
Hey, Daniel! I became curious about the negative database after reading your post, so I did some cyber-snooping and found a paper called "Protecting Data Privacy through Hard-to-Reverse Negative Databases." The idea seems to make sense (I must admit I'm drawing my conclusion based on the abstract, not the content of the entire paper): if you define sensitive data by what it is not, then the actual data itself is never explicitly revealed, so it can never be compromised.
Well, it seems to make sense only to an extent. What if, say, I'm searching through a database of single women in San Francisco for someone who reads Shakespeare, eats sushi, performs random acts of kindness to strangers, and smiles like Drew Barrymore? (It's a lot to ask, I know, but I'm constructing a theoretically argument here, so bear with me.) If the database were a negative one, I'll probably get a string of results showing everyone except the ones I'm looking for. In other words, I’ll be able to locate all the single women in San Francisco who don't read Shakespeare, don't eat sushi, don’t perform random acts of kindness, and don't smile like Drew Barrymore (maybe some of them smile like Jack Nicholson, which is unsettling).
Maybe a potential mate is not exactly what the scholars consider "sensitive data," so it doesn't fit their purpose.
Oh, wait, maybe the process of elimination--identifying everyone I shouldn't be looking for--is precisely what I need before I can find someone I'd be compatible with. Oh boy! This is getting pretty complicated, and I have yet to go on my first date!
Wow -- Kenneth Wong has some very interesting ideas and insights into the whole negative database issue! Certainly gave me a lot to think about -- thanks Kenneth!
wow - that's quite a response, K-dubs. I didn't realize someone was actually going to force me think about this for crissakes... the database you propose would be impossible I'd guess. To find your mate, you'd have to search for women who read Christopher Marlowe and smile like Lucy Lui. And every other possible combination of tastes and hobbies. For this kind of database to work, you'd have to have a finite set things that your data is NOT. And the article addresses that: "The important qualification concerns the word 'everything.' In practice, that means everything in a particular set of things."
So I suppose your idea for this negative dating site would be possible. It'd just be super-niche with a very small and discriminating clientele. Perhaps you and Ekman should start that site, hmm?
But seriously, I do like your new approach to dating--finding who it is you DON'T want will leave you with who you do want... right?
- daniel
Post a Comment