Artificial Intelligence Depot
Visiting guest. Why not sign in?
News, knowledge and discussion for the AI enthusiast.
FEATURES COMMUNITY KNOWLEDGE SEARCH  
Natural Language Understanding
Delivers a synthesis of major modern techniques in natural language processing. The approach is unique in its coverage of semantic interpretation and discourse alongside the foundational material in syntax.
More information at Amazon US UK
Randomisation for Web Privacy
IBM's Next Gen Data Mining
 
Randomisation for Web Privacy

The two biggest names in Data Mining, Rakesh Agrawal and Ramakrishnan Srikant, researchers at IBM, think they have come up with a solution for user privacy on the internet -- according to this article on The Register: IBM's Answer to Web Privacy. The trick is to add random values to the user's data when he submits it via a form. A data-mining algorithm, presumably tailored to deal with statistical analysis is then applied to retrieve accurate trends in the falsified data.

I have three issues with this technology... first, you need a lot of users to be able to get reasonable patterns when constant noise is applied. Secondly, shouldn't web privacy aim to allow the user not to enter his information at all; if not, why can't exact information be stored confidentially on a server, using a standard data-mining algorithm to extract trends? Thirdly, what stops a user from entering false data altogether! (a point raised in the article) Hmmm...

934 posts.
Tuesday 04 June, 14:32
Reply