WorldSciNet

In this paper, we propose an approach for preserving trade secrets in B2B interactions among competitors. Customer information exchanged during Business-to-Business (B2B) collaborations is usually considered as a business asset not to be freely shared with other businesses. This customer information is in essence a business trade secret. The full automation of B2B interactions is now possible because of the wide deployment of such technologies as Web services. However, these advances also create greater opportunities for businesses in acquiring sensitive customer data from other transacting businesses, thus creating an impediment for potential B2B collaboration between competitors. This calls for techniques to protect disclosure of sensitive data. Our approach leverages psycholinguistic knowledge to perturb data to computationally impede the disclosure of privileged customer information. We present an analytical model and a set of experiments to demonstrate the robustness of the proposed techniques.