Action Rules Discovery without Pre-existing Classification Rules

Action rules describe possible transitions of objects from one state to another with respect to a distinguished attribute. Previous research on action rule discovery usually requires the extraction of classification rules before constructing any action rule. In this paper, we present a new algorithm that discovers action rules directly from a decision system. It is a bottom-up strategy which has some similarity to systems ERID and LERS. Finally, it is shown how to manipulate the music score using action rules.