Interpreting association rules in granular data model via decision logic

Based on machine oriented modeling, a formal theory of association rules has been developed; the theory allow us to mining association rule mining by solving a set of linear inequality. Unfortunately, these rules are un-interpreted in the sense, we cannot generate a proper formula using the originally given symbols (attribute values) to described these discovered rules. In this paper, we develop a theory to generate such formula of given symbols to interpret them.