In many business and government systems it is necessary to index data about both the "good guys" and the "bad guys":

While the data stored may be identical, this is not a good reason for storing the information in the same file. If they are stored together and indexed together it is easy to miss a critical "bad guy".

In many system designs, a central name index, or personality file is created, with one common Name Search dialogue built for it. Then simply because it exists and contains names, addresses, account numbers, and other identity data together with system references, all forms of data are stored in this one "cross reference" index.

For both system performance and quality, and to allow user dialogues to be more efficient and effective, the records about negative or risk related information should be indexed separately using more exhaustive and expensive techniques for the negative data. Certainly the commonality of the process and formats can be taken advantage of by sharing code and inheriting designs, but mixing the good with the bad is never a strong design.

In order to maximize the chance of finding the high risk "bad guys" keep them in separate files, index them more exhaustively, and use wider search strategies.