Tuesday, July 29, 2008

complexity of anonymity

i was working on a problem to reduce the complexity of anonymising structured data, to ensure the privacy of individuals and achieved a factor of O(n) an improvement over the previous best solution which promised a complexity of O(nlogn)....(for all the non-geeks ... i somehow promised to solve this problem in a shorter time...:))......

So you must all be wondering , why the hell i am boring you to death by telling this....well coz all of us at some stage in our life have try to suppress various "key attributes" of ours ... so that we can hide our past. We dont want others to let know ...about the events which took place in our lives...may be we all are scared ..of what ???..well we should ask ourselves , if hiding your problems and difficulties can solve them , then i don't think there would be any in this world , sadly it's not the case......

The mask which we wear keeps on changing ... in front of our parents,relatives and our friends...we always try to project a cool image of ours... the "james bond" type ... but when this mask is revealed ... we stand in front of them ... naked and not able to look straight into their eyes....so we ignore them ....but for how long ???...... we all wish that we are able to suppress some critical information of ours so that others don't know about us .... we wish that "anonymizing" oneself is as easy as the solution that i somehow have been able to arrive in theory of mining...

But life's no dataset or a corpora that than be tailored and clustered to our wishes, we can't bulit training sets that function according to our commands. Training sets normally put data in data into two "portions" left or right ... but the opinions of people and their thinking can't be binary .... its between o and 1 .... and not 0 and 1.... no matter how long one try to cover one self with the veil .... this veil will reval our true face..... so it's better we dont try all this crap in our life's ...coz emerging out it can take ages... we may try to present a happy face... but deep inside we know how this can slowly eat our happiness .... so it's better to be your true to one self.....

So anonmyzing our key attributes can suppress some of our characteristics... but we forget ...that combination of our other attributes .... the so called "quasi identifiers"...in the language of database can reveal the facts we have tried to hide .....it's like a cat and dog race... i may have promised to solve the theoretical problem in O(n) time.... but some day , others will reduce the complexity further .... bringing out ways where this anonymity can be "unmasked"... and then proposing another way to anonymizing ourselves... but in the end we all are running a race , where we know that we will eventually loose..... so live as u want ... to live ..... GTH everyone......

PS: i may have used some jargons ...which are used in computer science theory ... but i hope everyone gets the point....

4 comments:

Phoenix said...

Why try so hard?

daman said...

[AT] phoenix
It's mot that i am trying hard ... but others try hard to fake themselves.. but then thry dont realise that ...one day or another things would be revealed.....

Anonymous said...

As long as the people who matter and you know the truth, I'd say rest of them all should GTH!

Btw, nice blog you've got here.... a nice CS angle to life (being a CS major myself.... I appreciate it a lot!) keep it going!

Suri said...

u were right..

i really did not get the first para :)