Nicholas Carr points out this link: Data Mining 101: Finding Subversives with Amazon Wishlists, a very good but disconcerting example of what can be stitched (mashed) together about you. Tom Owad used data from a random sampling of Amazon wishlists to create a profile of a person's interests, such as reading subversive books, that some government agencies might be interested in. He then then aggregated this data and plotted the exact location of those people on a map.
The demonstration follows these steps:
- Searches on the name "Edgar", in homage to J. Edgar Hoover, and finds 260,000 wishlists.
- Aggregates on titles like "Fahrenheit 451", "Build Your Own Laser", and "Koran", which returns a healthy sample.
- The Ontok Geocoder service is used to convert the wishlist addressee information to latitude and longitude coordinates.
- Then finally, those addresses are accurately plotted on a Google Map.
The image on the right shows readers of "1984".
Well done, clever and a bit scary. Careful what you wish for indeed.