I am having a problem trying to solve a question on a past paper asking to design a security lattice. Here is the question:
The AB model (Almost Biba) is a model for expressing integrity policies rather than confidentiality. It has the same setup as Bell-LaPadula, except that $L$ is now a set of integrity levels which express the degree of confidence we have in the integrity of subjects and objects. Subjects and data at higher integrity levels are considered to be more accurate or safe. The set of subjects and objects may also be different, for example, programs are naturally considered as subjects.
Often, the set $L$ is actually a lattice of levels, with two operations: least upper bound $l_1 \vee l_2$ and greatest lower bound $l_1 \wedge l_2$, where $l_1, l_2 \in L$.
i. Design an example integrity lattice for AB, by combining two degrees of data integrity dirty and clean and two means by which a piece of input may be received, website (external user input from a web site form) and dataentry (internal user input by trusted staff).
I have been looking for an explanation on how to build lattices but can't seem to find one on the internet or in textbooks. Can anyone point me in the right direction?