Comment by zqfm

Maybe off the wall, but my first thought is certificate revocation. Could there be (or is there already) a semi-centralized database of research that should no longer be cited? With maybe a dependency graph of research that maybe needs to be revisited?