A vertically integrated market consists of potentially competing companies who supply each other with products and services. Information technology can be employed to improve the business transactions and the flow of information within the market. In this paper, we show a solution on integrating massively heterogeneous product catalogs into multi-language and multi-role product groups organized as ontologies. Besides product classification, the method also addresses the classification of product properties. The product catalog data structure is decomposed into a set of label-value pairs which are then classified into the multiple ontologies which is the basis for query formulation. Our approach has been realised within the European research project MEMO (Mediating and Monitoring Electronic Commerce) using the construction industry in an extensive case study. The method can also be applied to related areas like multi-media databases or digital libraries.
Manfred A. Jeusfeld