Building an automatic e-tendering system on the Semantic Web

Over the years, business-to-business (B2B) e-commerce applications have been well researched, but the degree of automation achieved has been limited. This is probably because the current content of the Web is designed for human comprehension, rather than for computer operations. In B2B, an e-tendering system needs to contact a large number of potential sellers and negotiate deals with them. It would be helpful if this process could be carried out automatically. This study proposes an automatic e-tendering system that implements an automatic negotiation process over the Semantic Web in which Web pages provide information not only through their content, but also through the properties of that content. The e-tendering system integrates a negotiation process that considers bargaining power and risk preference of negotiators. The semantic data then are used to select tenders or to determine negotiation strategies. A system is built in a peer-to-peer (P2P) environment to simulate a two-player negotiation process for demonstration purposes. The system can automatically derive negotiation results for different risk preferences (such as risk neutrality, risk aversion, or risk proneness) and degrees of negotiating power (such as the cost of haggling).