Measuring the Quality of Descriptive Languages for Products and Services

Abstract

Standardized product classification systems play a major role for automated business transactions. They are not only a coding scheme that eases catalog data integration, but can act as descriptive languages for products and services, supporting a multiplicity of future e-business scenarios. Even though prominent approaches like eCl@ss and eOTD are built around a hierarchical order of product and service categories, this is not a compulsory property. The key functionality is the representation of business meanings (e.g. a product or service) in an unambiguous, machine processable manner. Currently, the most popular architecture for such descriptive languages is the combination of three components: classes, an attribute library, and class-specific attribute lists, which consist of references to characteristic attributes for the respective class. The development and maintenance of the respective attribute lists is a major challenge, as this requires comprehensive domain knowledge about the respective goods. Eventually, the quality and usefulness of the descriptive language is determined by the quality of its attribute lists. This paper proposes metrics that help measure the quality of and the progress in the development of descriptive languages for products and services.

Keywords: Product classification, eCl@ss, eOTD, UNSPSC, attribute lists

Keywords: Product classification, eCl@ss, eOTD, UNSPSC, attribute lists, Semantic Web, Ontologies, Content Quality

JEL Classification: Z00, L15