In this paper, we introduce association rules into the event matching process, and propose a matching algorithm based on association rules in ontologybased publish/subscribe system. The algorithm discovers association rules from subscriptions, and then integrates these association rules into the arrived events. These integrated association rules guide the event matching process, which can greatly improve efficiency by reducing unnecessary matching processes. To solve the available problem of integrating the association rules into the ontology-based publish/subscribe system, we also show an approach of transforming the subscriptions into a format that can apply the data mining algorithm and an approach of integrating the association rule into the events. The results show that the algorithm achieves performance improvement of matching efficiency by using comparison.