A method for generating a categorization scheme including a set of one or more categories for a collection of data items, each data item including at least one data element having an element name and an element value, the method comprising the steps of: identifying a common data element for each one of the data items, each of the identified common data elements having equivalent element names; compiling a collection of element values comprising an element value for each of the common data elements; and defining the set of one or more categories to contain the different values in the collection of element values.