The E-Commerce Product Classification Challenge

Authors:
Ellen Schulten, Heloise Ontology Associates
Hans Akkermans, Free University Amsterdam
Nicola Guarino, LADSEB
Guy Botquin, Content Europe
Nelson Lopes, Institut der Deutschen Wirtschaft Köln
Martin Dörr, ICS-FORTH
Norman Sadeh, Carnegie-Mellon

Resume

This article launches an international research challenge in the area of intelligent e-business. The challenge is to come up with a generic model and working solution that is able to (semi-)automatically map a given product description between two different e-commerce product classification standards.

B2B E-Commerce: in search for a Lingua Franca or Rosetta Stone

A fundamental premise - and a major economic driver - behind B2B Electronic Commerce is that labor-intensive and time-consuming human interactions can be replaced with (semi-)automated Internet-enabled processes. Looking at actual implementations, we indeed see ‘simple’ applications, such as product search and
selection without the intervention of a sales representative, and more sophisticated solutions, such as server-to-server communication for inventory replenishment of enterprises.
Nevertheless, the hard reality of e-business, and the slower than (some industry watchers) expected adoption of electronic buying, points to the complexity of replacing human interactions by computer. Of course this is not difficult to understand. In the human world, dialog is structured by grammatical, semantic, and syntactic rules that live in a shared context of social and cultural conventions. The young e-commerce world still
lacks this rich background, and we are still far from really achieving the vision of a ‘Universe of Network Accessible Information’ - as the Web is defined by the W3C. The existing protocol and hypertext standards appear to be insufficient for the exchange of unambiguous information between network devices - an essential condition for electronic commerce. This is especially true in business between enterprises, where interoperability between different systems, back-end system integration, complex business processes, and
stringent business rules have to be taken into account.

Fonte:http://www-2.cs.cmu.edu/~sadeh/Publications/More%20Complete%20List/OntoWeb%20Challenge.pdf