On Trustworthy Rule-Based Models and Explanations