One project I always wanted to do is create a publicly searchable database which mined connections between public officials and their financial interests.<p>It would kind of be like a Facebook which the users had no control over their profiles. All known associates of an individual would be linked together in a giant social network. An algorithm would flag conflicts of interest. This should make it easy to see problems with connections a step away from a public official rather than just high profile associates or close family members.<p>Automated alerts would trigger when individuals within a public official's group are arrested, convicted of a crime, etc. Much like Google's algorithm identifies "bad neighborhoods" of web site links, you could instantly see if a particular public official was surrounded by questionable or explicitly corrupt characters.<p>Most importantly this could be used to generate an easy to understand corruption score, making it simple for the public at large to understand how corrupt the individual is. Once a particular politician or official scored past a certain point they may become an outcast, finding potential campaign fundraisers and advocates avoiding them.<p>The migration of eyeballs away from television and old media is going to have a profound impact on the democratic political space. Imagine if a political candidates advertisements delivered via Google also contained a corruption score next to them, much like how e-commerce stores on Adwords display user ratings.