Smishing is a cyber-security attack, which utilizes Short Message Service (SMS) to steal personal credentials of mobile users. The trust level of users on their smart devices has attracted attackers for performing various mobile security attacks like Smishing. In this paper, we implement the rule-based data mining classification approach in the detection of smishing messages. The proposed approach identified nine rules which can efficiently filter smishing SMS from the genuine one. Further, our approach applies rule-based classification algorithms to train these outstanding rules. Since the SMS text messages are very short and generally written in Lingo language, we have used text normalization to convert them into standard form to obtain better rules. The performance of the proposed approach is evaluated, and it achieved more than 99% true negative rate. Furthermore, the proposed approach is very efficient for the detection of the zero hour attack too.