Abstract
The useful knowledge exists in a large amount of data and we extracted fuzzy rules from the data of given attributes in the conventional method. In this paper, we propose a method to extract fuzzy rules including new attributes generated by the arithmetic operations of attributes. We generate a simple data set where a symbolic class has a string association with the multiplication of two numerical attributes. We extract fuzzy rules between two numerical attributes from the data set with the conventional method and examine the associations between two attributes. Consequently we can select a pair of attributes of a middle degree of associations to calculate by four arithmetic operations to generate new attributes. We add them to the original attributes to extract fuzzy rules with the conventional method. The result shows that it extracts rules including new attributes of the multiplication of selected attributes.
