Comment by matte_black

Cleanup is standard work for any dataset. With enough money and a pile of Indians working for cheap hourly rates you can clean through A LOT of data. 10,000? Try a million!