Wikipedia: The Text Adventure by Kevan Davis.
You can choose a starting location or enter any Wikipedia article as your starting point. Described by Davis as interactive non-fiction.
Interesting way to learn an area but the graphics leave much to be desired.
If you like games, see http://kevan.org/ for a number of games designed by Davis.
Paparazzi, for example, with some modifications, could be adapted to being a data mining exercise with public data feeds. “Public” in the sense the data feeds are being obtained from public cameras.
Images from congressional hearings for example. All of the people in those images, aside from the members of congress and the witness, have identities and possibly access to information of interest to you. The same is true for people observed going to and from federal or state government offices.
Crowd-sourcing identification of people in such images, assuming you have pre-clustered them by image similarity, could make government staff and visitors more transparent than they are at present.
Enjoy the Wikipedia text adventure and mine the list of games for ideas on building data-related games.