Classification
- Tags
- security
We define [see page 10, classification] as constructing a hierarchy of classes of security levels. eg. Top secret must be more closely protected than restricted or unclassified.
The classes we commonly use in order from highest-to-lowest sensitivity are:
1. Top_Secret
2. Secret
3. Confidential
4. Restricted
5. Unclassified
Note: Security classes often form a [see page 43, lattice].
Mathematical Properties
The relationships between classes is defined as see page 11, dominant, and has the following properties:
Term | Description |
---|---|
[see page 12, Reflexive] | Each class \( x \) dominates (is equivalent to) itself. |
[see page 13, Asymmetric] | If \( x \) dominates \( y \) then \( y \) can never dominate \( x \). |
[see page 14, Transitivity] | If \( x \) dominates \( y \) and \( y \) dominates \( z \), then \( x \) dominates \( z \). |
Caveats and Codewords
We can also further expand classification [see page 40, into]:
- caveats, restrict access based on class, such as UK only (or Canadian+US+UK).
- codewords (see codebook), keywords used to obscure information on a need-to-know basis.