Data mining is widely regarded as Evil. Our purpose here is to explore the possibility of data mining as an appropriate technology.
Let's start with the observation that individuals (as contrasted with "institutions") possess large amounts of actionable data. One of the main barriers to mining these data is their not being machine readable.
The traditional means for enlarging the machine readable domain is an activity called "data entry." This used to be an occupation in the United States. Now, of course, it is the most trivially easy service to export, so opportunities to get paid for it are largely a thing of the past. This may work to the advantage of public domain data miners, as an unemployed data entry operator has the ideal skill set for such activity, minus the conflict of interest inherent in employed status in the informational sector.
Everyone, of course, is welcome to volunteer in such a movement.
A similar (according to some) concept is knowledge management.