A combinatorial approach to information inequalities

The authors present theorems that lead to a new combinatorial approach to prove information inequality. The non-negativity of Shannon's information measures can all be derived using this combinatorial approach, and each of these inequalities has a very simple intuitive combinatorial explanation.