A typical Hamiltonian for Ising model is $$ H=-\sum_{i,j} J_{ij}S_iS_j - K \sum_i S_i.$$ In many references we can find exact solutions for special cases, mean-field approach, phase transition, and many others. However, I cannot find how valid is this model in real physical materials, and what is the microscopic origin of the phenomenological constants $J$ and $K$.
What is the microscopic origin of this Hamiltonian? Can this Hamiltonian be obtained by a more systematic approximation of a more general Hamiltonian that describes magnetic systems?
For what type of material this model is valid? Since there are enormous types of interactions that can occur in solids and Ising model features only some of them, in what cases the "Ising-interaction" is dominant than others?