問題文
A classification scheme has four levels but engineers routinely mark new data sets at the second level regardless of content. What change to the scheme is most likely to improve the accuracy of classification?
選択肢
- Define each level by the concrete handling rules it triggers, so the choice of level has visible consequences the engineer can reason about.
- Reduce the number of levels from four to two so that the decision is simpler, since a binary choice between sensitive and non-sensitive eliminates the ambiguity that causes engineers to default to the middle option.
- Require a second person to approve the classification of every new data set.
- Change the default value in the registration form to the highest level.