Thursday, 21 April 2016

Staying private while searching the Internet (by Alessandro Checco)

One of my interests in online data privacy: how can users access custom content without being tracked individually? Can we break this vicious circle in which advertisers spy on users and, as a reaction, users hide even more?

My idea is to allow a milder form of identification than the classic way in which a use is uniquely identified.  Instead, users would automatically be hidden within a crowd of similar users.

The challenges of this approach is to combat spam and Sybil attacks, but it turns out it can be easily done through cryptographic tokens such as e-cash.

Another topic I am exploring is: how to detect Search Engines ‘learning’ about sensitive topics during a searching session? Interestingly, the advertisements that appear during searches provide evidence of tracking on sensitive topics. Google doesn't seem to track our entire search histories for the purpose of advertising, but just the last 4-5 queries.