"For example, we see the peaks and valleys in problem arrival rates in one stream and map them to events such as verification test runs, field trials, etc." . . . .