In e-procurement, companies use e-catalogues to exchange product infor-mation with business partners. Matching e-catalogues with product requests helps the suppliers to identify the best business opportunities in B2B e-Marketplaces. But various ways to specify products and the large variety of e-catalogue formats used by different business actors makes it difficult. This Ph.D. thesis aims to discover potential syntactic and semantic rela-tionships among product data in procurement documents and exploit it to find similar e-catalogues. Using a Concept-based Vector Space Model, product data and its semantic interpretation is used to find the correlation of product data. In order to identify important terms in procurement documents, standard e-catalogues and e-tenders are used as a resource to train a Product Named Entity Recognizer to find B2B product mentions in e-catalogues. The proposed approach makes it possible to use the benefits of all availa-ble semantic resources and schemas but not to be dependent on any specific as-sumption. The solution can serve as a B2B product search system in e-Procurement platforms and e-Marketplaces.