Research on extracting rules from information system based on discernible vector

The problem of rule extraction is an important part of knowledge database discovery, so the paper proposed a new method to deal with it. Established a discernible vector based on equivalent class relationship, operated the discernible attributes in the discernible vectors through boolean calculation, and then obtain the reduced attribute set. Finally, the concise decision rules were obtained after reducing the attribute values in the reduced attribute set. The illustration and experiment results indicate that the method is effective and efficient for rule extraction.