Operations for granular computing: mixing words and numbers
Ronald R. Yager, Dimitar Filev · 2002
We discuss the idea of granular computing and indicate its importance to the agenda of intelligence engineering. It is noted that attributes of granular objects are usually measured on a very weak scale, often a linguistic scale having only a linear ordering. The need for the operations to manipulate granular objects is presented. We then introduce a new form of the ordered weighted averaging (OWA) operator, called the induced OWA (IOWA) operator, in which the ordering of the objects to be aggregated can be based on something other then the values to be aggregated. We then show that this new IOWA operator provides a tool in which we can perform operations which mix numbers and words.