Jeff Jonas, a top practitioner in data analysis, has criticized one type of data-mining known as "Non-Obvious Relationship Awareness" as a total waste of money and time. (The Pentagon's Total Information Awareness supposedly incorporated this, as well as the more conventional, form of mining.) The idea here is to collect all sorts of data about terrorists' traits and behavior—and then scan similar databanks for the entire population to find people with similar traits and behavior on the theory that they might be terrorists, too. Jonas' argument is that such matches are entirely spurious and so should not serve as even the starting point of any "terrorist search."