error seeding. (IEEE) The process of intentionally adding known faults to those already in a computer program for the purpose of monitoring the rate of detection and removal, and estimating the number of faults remaining in the program. Contrast with mutation analysis.