Introduction of entropy functions to improve the classification of commodities in e-commerce

In classification systems like eCl@ss the number of classes grows and products like for example, screws are spread across several segments related to selected applications or industries. Entropy functions on test sets are used to calculate the dispersion of products across the classification system. For such dispersed classes the corresponding property sets are compared and it is shown that there is a real need for the harmonisation of property sets to improve e-commerce applications.