In essence, this involves a sophisticated mathematical estimate of the change in probability added by the new evidence, obtained by contrasting ‘prior probability’ (before taking into account the new evidence) with ‘posterior probability’ (after taking the evidence into account).
