Automated negotiation is the key techniques for reaching agreements in agent-mediated electronic commerce. Current automated negotiation models assume that users know the value of the product or service they want to buy and provide their agents with a reservation price, and the agents make offers and reach agreements with other agents according to this reservation price. However, in real world electronic marketplaces users probably do not know the exact value of the item, which is in terms of price, and the reservation price they set to their agents only means the maximum price they are willing to pay for the item. In this paper, we propose a negotiation model to deal with the valuation problem. The shopping agent in our model can deliberate the market price and the seller agent’s reservation price from public available information and the seller agent’s proposals. Also in order to conform with the real world negotiation conditions, we introduce negotiation features of real world human shopping such as multiple sellers, valuation, ultimatum, and learning from available information, etc. into our model.
展开▼
