Matching heterogeneous e-catalogues in B2B marketplaces using vector space model

Abstract

Competitive pressures require manufacturers to continually find ways to reduce costs and increase market access. Internet-based technologies such as e-procurement (electronic procurement) provide increased buying and selling opportunities to manufacturers. E-catalogues are electronic representations of information about the products and services of an enterprise and play a key role in e-procurement. Matching an e-catalogue from a buyer with product e-catalogues provided by the suppliers helps companies to find partners in e-marketplaces. Since each business actor may use a different structure and classification for e-catalogues, it is not easy to match a product with the e-catalogue requested by another partner. This article uses vector space model (VSM) and customises it to solve the matching problem of the e-catalogues. The proposed approach uses a combinations of values, names and location of attributes of structured documents to find the syntactic correlation of e-catalogues. It also uses ontologies to expand the matching mechanism with semantic relationships of data attributes. The proposed approach makes it possible to use the benefits of all available ontologies and schemas but not to be dependent on them. The experimental results show that the proposed approach provides a more accurate similarity ratio while matching similar e-catalogues compared with the basic approach of just using VSM.