In this paper, an intelligent approach for B2B electronic commerce to help businesses choose their negotiating parties is presented. The premise is that the participants are often less concerned with price and more with relationships in B2B commerce. Our approach, as proposed in this paper, consists of two main steps. First, a social network analysis method is used to mine the participant relationships on market space automatically. Second, a neural network method is used to recommend the most suitable trading partners to a user in the online marketplace before implementing actual negotiations. Relationships mined in the first step will be used as important parameters in the proposed neural network approach. The resulted neural network model is used by each participant to estimate the negotiation results with its potential negotiation partners. The intelligent approach is tested on the B2B platform we developed. It is quite useful to diminish the digital gaps among the participants and save energies in actual negotiation by using the estimated negotiation results. In addition, it is a helpful supporting tool for newcomers of the platform to enter the market. Furthermore, we provide an interface for users to set their preferred weights of relationships, prices, and the cost of negotiations, and applying their preferences to the trading partner selection in actual negotiations.
